Spestcsbeseelie een icbetetuaal we negera weevewrs tomes wept, Hay hu mh ey , ma aN Digitized by the Internet Archive in 2009 with funding from Ontario Council of University Libraries http://www.archive.org/details/proceedingscalif01 calli i ic : (ey yy i ih } f) Hei Tl) Wh Bn Nndaeent ie en i { + ea ea Nh i AN 10 | ae aa ul we Ai BA a PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FOURTH SERIES Vou. I 1907-1912 BAG oa fe SAN FRANCISCO PUBLISHED BY THE ACADEMY 1912 A ty, qt My Oy, + et ~( { P . f; { aed - ee Vie Js eee ; a M\\ me aon ‘ © CONTENTS OF VOLUME I. Piates I-XXXIV. PAGE Mithe-papemers neat ecto cre watt ns Dee en ANB, Ante DE LEER Hae eet i (CRESTS TRS) Ba ceere cee ores eee error te een RE OT Es RS EN OF LP Peed ili No. I. Preliminary Description of Four New Races of Gigantic Land Tortoises from the Galapagos Islands. By John Wan enbungint) soccer Ca erent eth 2 Oke eee re ee ae 1 (Published December 20, 1907) No. If. A Botanical Survey of the Galapagos Islands. By Alban Stewarts (@Platesmd— Kuler se snes eens 7 (Published January 20, 1911) No. III. The Butterflies and Hawk-Moths of the Galapagos Islands. By Francis X. Williams. (Plates XX-XXI).. 289 (Published October 7, 1911) No. IV. The Snakes of the Galapagos Islands. By John Van Denburghy W@Blates: MUL ONENG ee seceesecceceesccetecnstcen 323 (Published January 17, 1912) No. V. Notes on the Botany of Cocos Island. By Alban Stewart. (@Blatesy XEXCSTXeXONATNV A) Re eee BR ress cent 375 (Published January 19, 1912) No. VI. The Geckos of the Galapagos Archipelago. By John Van Mera bar gla ee eae pen ee ge nner Pt ot 405 (Published April 16, 1912) No. VII. Notes on the Lichens of the Galapagos Islands. By Alban SLE Wal Eta cre se rekon SRI SL Se LS oe ates 431 December 30, 1914. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES FourtTH SERIES Vot. I, pp. 1-6. December 20, 1907. EXPEDITION OF THE CALIFORNIA ACADEMY OF SCIENCES TO THE GALAPAGOS ISLANDS, 1905-1906. I. PRELIMINARY DESCRIPTIONS OF FOUR NEW RACES OF GIGANTIC LAND TORTOISES FROM THE GALAPAGOS ISLANDS. BY JOHN VAN DENBURGH, Curator of the Department of Herpetology. Early in 1905 the California Academy of Sciences decided to send an expedition to the Galapagos Islands. The general purpose was to explore this group more thoroughly than the opportunities of previous investigators had permitted, and to secure large collections of the plants, mollusks, insects, birds, mammals, and reptiles in the hope of throwing more definite light upon the origin of the archipelago. Particularly, it was determined to study the geology of the islands, to make a very careful search for fossils, and to spare no effort to secure specimens or remains of those races of the gigantic land tortoises which long had been thought extinct. Study of the published results of previous expeditions had convinced me not only that these islands must all, at some December 19, 1907, 2 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. former period, have been parts of a single land-mass, becom- ing later, by partial submersion, separated into the various islands, but that Albemarle Island, which possesses several races of tortoises and on which Heller and Snodgrass found evidence of an elevation amounting to several hundred feet, had much more recently been formed by the union of several smaller islands corresponding, probably, to its five great vol- canoes. Accordingly, the members of the expedition were instructed to collect on Albemarle exactly as though it still were five islands. The expedition set sail from San Francisco, on the twenty- eighth of June, 1905, in the schooner “Academy,” which had been purchased and rechristened for the purpose. The scientific staff of the expedition consisted of eight young men. Mr. R. H. Beck, who has had more experience in these islands than any other collector, this being his fourth expe- dition to them, was in charge. Mr. Alban Stewart went as botanist; Mr. W. H. Ochsner, as geologist; Mr. F. X. Will- iams, as entomologist; while Mr. E. W. Gifford and Mr. J. S. Hunter were to study and collect the birds, and my assistant Mr. J. R. Slevin, with the aid of Mr. E. S. King, was to care for the reptiles. Having made brief stops at various islands near the coast of Lower California, as well as at San Benedicto, Socorro, Clipperton, and Cocos Islands, the party reached the Galapagos Archipelago and landed upon Hood Island, September 24, 1905. During the months which followed the most arduous collecting was vigorously carried on in all the islands of the group, many of the larger being visited several times, and on September 25, 1906, after a full year of work, the “Academy” left Culpepper Island and set sail for San Fran- cisco, where she arrived in safety Thanksgiving Day, Novem- ber 29, 1906. The collections brought back are by far the largest and most important ever gathered in these islands. The reptiles num- ber over forty-five hundred specimens, of which nearly four thousand are from the Galapagos. The search for land tor- Vot. I] VAN DENBURGH—GIGANTIC LAND TORTOISES 3 toises met with far greater success than I had dared anticipate. All of the races which had been supposed extinct were found still living, with the exception of that of Charles Island. Tor- toises were also found living on two islands which they had not previously been known to inhabit. On Barrington Island, also not hitherto known to have supported tortoises, por- tions of the remains of fourteen individuals were secured. It is probable that the tortoise of this island, like that of Charles, is really extinct. A single tortoise was secured on Cowley Mountain, Albemarle Island, and others were found living in all of the other localities from which these huge reptiles have ever been recorded. In all over three hundred tortoises are represented in the collection, some forty of them, however, only by more or less fragmentary remains. A complete report upon this collection can only be issued after an immense amount of work. Meanwhile, it seems best to publish this brief statement and the following pre- liminary descriptions of the tortoises of Hood, James, Chat- ham, and Narborough Islands, which seem never to have been described. Testudo hoodensis new species. Type—Adult (?) female (?) now living in Golden Gate Park, San Francisco. California Academy of Sciences No, 8121. Hood Island, Galapagos Archipelago. Joseph R. Slevin and E. S. King. Caught June 27, 1906. Diagnosis—No nuchal; gulars paired; front of carapace high, little lower than middle, height at nuchal notch more than 41% (45%) of straight length; difference between percentages of heights at third vertebral and at nuchal notch less than 9 (5); carapace saddle-shaped, narrow anteriorly, width at margin of junction of second and third marginals not more than 54% (45%); first marginals not greatly enlarged, not much everted, their ventral surfaces not vertical, their most prominent points separated by less than 30% (20%); length over curve not more than 123% (122%), greater than width over curve; vertical distance from lower surface of plastron to lower edge of lateral marginals great, 12%; general size rather small, straight length (June, 1907) 22.2 inches; plastron long, median length 89%; plates striated, central portions of vertebrals and costals much elevated; pectoral plates forming a suture on median line; lower jaw and throat marked with yellow. 4 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. Testudo darwini new species. Type—Adult male. California Academy of Sciences No. 8108. James Island, Galapagos Archipelago. R. H. Beck and Joseph R. Slevin. July 31, 1906. Diagnosis—No nuchal; gulars paired; fourth cervical vertebra biconvex;' carapace high, elongate, somewhat dome-shaped but high in front; posterior declivity beginning about middle of third verte- bral; height at nuchal notch more than 41% (45%) of straight length; difference between percentages of height at third vertebral and at nuchal notch more than 9 (10); carapace not saddle-shaped, width at margin of junction of second and third marginals 55%; width over curve in male not greater than length over curve; vertical distance from lower surface of plastron to lower edge of lateral marginals moderately great (9%); general size large, straight length 38 inches; shell heavy; pectoral plates forming a suture on median line; the sum of the measurements of the length over curve, length of plastron, height at nuchal notch, and height at third vertebral, equals or exceeds the sum of the measurements of the straight length, straight width, and width over curve; jaws and throat black. Testudo chathamensis new species. Type—Skeleton of adult male. California Academy of Sciences No. 8127. Found in a cave on Chatham Island, Galapagos Archipelago. R. H. Beck and Joseph R. Slevin. February 12-14, 1906. Diagnosis—No nuchal; gulars paired; fourth cervical vertebra biconvex; carapace depressed, front elevated in male; height at nuchal notch less than 41% of straight length (male 34, female 27%); male flat-backed, female dome-shaped, difference between percentages of heights at third vertebral and at nuchal notch 6 in male, 24 in female; carapace of male slightly saddle-shaped but broad, width at margin of junction of second and third marginals 53% in male; anterior marginals but little everted; length over curve in male 112%, female 126%; vertical distance from lower surface of plastron to lower edge of lateral marginals small, 4% in male, 6% in female; general size moderate, straight length in male 35.25 inches, female 22.5 inches; pectoral plates much reduced, not meeting on mid-line; jaws and throat of female black. Testudo phantasticus new species. Type—Adult male. California Academy of Sciences No. 8101. Nar- borough Island, Galapagos Archipelago. R. H. Beck. April 5, 1906. Mt is probable that this is the normal arrangement in all the races of the Galapagos Islands. The third cervical vertebra has been found biconvex in the types of 7. galapago- ensis and T. decki; but in seven specimens of the latter the fourth is biconvex, as in other races. Vot. I] VAN DENBURGH—GIGANTIC LAND TORTOISES 5 Diagnosis—No nuchal; gulars paired; fourth cervical vertebra biconvex; front of carapace high, not lower than middle, height at nuchal notch more than 41% (54%) of straight length; difference between percentages of height at third vertebral and at nuchal notch less than 9 (2); carapace saddle-shaped, narrow anteriorly, width at margin of junction of second and third marginals not more than 54% (46%); first marginals much enlarged, everted more than in any other race, their ventral surfaces nearly vertical, their edges from nuchal notch to prominent point nearly horizontal, prominent point almost a right angle; distance between prominent points of first marginals more than 30% (32%); length over curve more than 123% (124%), greater than width over curve; vertical distance from lower surface of plastron to lower edge of lateral marginals small, 6%; general size moderate, straight length 34.5 inches; plastron short, 70%; pectoral plates forming a suture on median line; lower jaw and throat marked with yellow. A few words in explanation of the measurements given in the foregoing descriptions may be necessary. In attempting to avoid the indefiniteness which has too largely characterized descriptions of these tortoises it was quickly found necessary to devise some means of expressing and comparing upon paper their individual variation in shape. This, it was found, could best be done by taking numerous measurements of each tor- toise and reducing all these measurements to percentages of the (straight) length of the tortoise. In this way, the measure- ments of tortoises of all sizes may be directly compared. The tortoise is placed upon a level board or table in such a position that it rests naturally upon, as nearly as possible, the entire length of the plastral bridge of each side. With the tortoise in this position, the straight length is the distance between verticals erected at the nuchal notch and at the posterior border of the supracaudal plate. The straight width is the distance between verticals erected at the sides of the tortoise opposite the line of meeting of the second and third costal plates. The curved length is measured with a tape-measure over the midvertebral line from the nuchal notch to the pos- terior edge of the supracaudal plate. The curved width is taken from the bend in the marginal plates up along the line of meeting of the second and third costals, across the middle of the third vertebral, down between the second and third costals, to the line of bending of the marginals. The width 6 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER, second to third marginals is the straight width at the level of the lateral margins of the sutures between the second and third marginal plates of each side. The middle height is the vertical distance between the board or table and the middle of the third vertebral plate, and is taken with a square and spirit-level. The front height is taken in the same manner at the nuchal notch. The height to marginals is the vertical distance from the table to the lower border of the marginal plates at about the middle of the plastral bridge. The plastron is measured with a tape along the median line, the tape is not pushed into plastral depressions and when the plastron is notched the projections are not measured. SAN FRANCISCO, November 18, 1907. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FourtH SERIES Vot. I, pp. 7-288 January 20, 1911 EXPEDITION OF THE CALIFORNIA ACADEMY OF SCIENCES TO THE GALAPAGOS ISLANDS, 1905-1906 II A BOTANICAL SURVEY OF THE GALAPAGOS ISLANDS BY ALBAN STEWART Botanist to the Expedition CONTENTS Pirates I-XIX PAGE INTRODUCTION : : ; ; : : : F : A F 8 ACCOUNT OF THE SPECIES OF VASCULAR PLANTS f é : A 11 BoTANICAL REGIONS . é ' ; j . ‘ ; ; § 206 GENERAL FEATURES OF THE FLORA. , : j ‘ : 5 211 EcoLtocicaL Factors. ; j j : : , R : : 219 OrIGIN oF THE GALAPAGOS ISLANDS . , 5 ; ; F ; 233 ORIGIN OF THE FLora . j 4 : A i : } : : 239 BIBLIOGRAPHY OF THE BoTANY OF THE GALAPAGOS ISLANDS. : 246 INDEX . i c : ‘ : : j : ; 4 : : 249 EXPLANATION OF PLATES. 5 , : 5 i : 4 : 254 January 14, 1911 8 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H SER. INTRODUCTION In the spring of 1905 I received the appointment of botanist of the scientific expedition sent to the Galapagos Islands by the California Academy of Sciences. In preparing for this expedition the California Academy purchased the U. S. Ship “Ernest,” a two masted schooner of eighty-seven tons burden, and after refitting, rechristened her the “Academy.” Our party consisted of eleven members, as follows: R. H. Beck, chief; F. X. Williams, entomologist; W. H. Ochsner, geologist and conchologist; J. R. Slevin, herpetologist; J. S. Hunter and E. W. Gifford, ornithologists; E. S. King, assistant herpetologist; Frederick T. Nelson, mate; J. J. Parker, navigator; James W. White, cook; and myself, botanist. All of the scientific members of the expedition shipped as seamen, so that the expedition was made up mostly of sailor-scientists. The expedition left San Francisco on the morning of June 28, 1905, and arrived at Hood Island, the most southern member of the Galapagos group, on September 24, nearly three months having been consumed on the trip, during which short stops were made at Ensenada, Lower California, and on San Martin, San Benito, San Geronimo, Cerros, Natividad, San Benedicto, Socorro, and Clipperton islands, Mexico, and Cocos Island, Costa Rica, on the most of which small collec- tions of plants were made. The expedition left the Galapagos Islands on the 25th of the following September, so that a year and one day was spent in the archipelago, during which time all of the islands were visited at least once, and the larger and more important ones two or more times at different seasons of the year. Up to the present time our knowledge of the flora of the Galapagos Islands has been due mainly to the collections of Darwin, Andersson, Baur, and Snodgrass and Heller, and to the writings of Hooker, Andersson, and Robinson.' 1 For a table of the botanical collections made on the Galapagos Islands, see Robin- son, Flora of the Galapagos Islands, Proceedings of the American Academy of Arts and Sciences, v. 38, no. 4, pp. 221-223. Vor. I) STEWART—BOTANY OF THE GALAPAGOS ISLANDS 9 Unfortunately many of the former collectors of plants remained but a short time upon the islands, and as most of them were not botanists, our knowledge of the general botanical conditions has remained rather meager. It was the intention at first to incorporate the entire botani- cal results of the expedition in a single paper, but as the present paper has assumed greater proportions than was expected, it seems best to divide the subject and publish the parts separately. The present paper consists of a rather detailed account of the different species of vascular plants, including their range in elevation and their distribution on the different islands; a brief description of the different botani- cal regions; an account of the general features of the flora; an account of the factors governing the growth of vegetation; and an account, so far as possible, of the evidence offered by the collection concerning the origin of the islands and of the flora. A second paper will deal entirely with a description of the botanical conditions on each island of the group, and short papers will treat of the lichens and mosses. The plan of treatment of the first part of this paper is in general the same as that pursued by Robinson in his “Flora of the Galapagos Islands,” as I was unable to devise a plan which I thought would be better. Many of the statistical tables are simply revisions of the tables as given by Dr. Robinson, although a few new ones have been added where it seemed necessary. The entire nomenclature has been carefully gone over and revised to make it conform with the new rules of the Vienna conference. The ferns have been treated as a single family and not split up into several different families as has been done by some authors. Treating the group in this manner has enabled me to handle it to better advantage in the latter part of this paper. With but one exception the nomenclature of Christensen, “Index Filicum,” has been used in this family. Unfortunately none of our instruments of measurement were graduated with the metric system, other- wise it would have been used. In order to economize space, the names of former collectors are only mentioned where there 10 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. are neither specimens or notes of a species from a given locality in the collection under consideration. The collections of vascular plants were identified by myself at the Gray Herbarium of Harvard University under the direction of Dr. B. L. Robinson, Curator of the Gray Herbarium. I wish here to express my thanks to Dr. Robinson for his kindness in giving me free access to the excellent collections of plants from the Galapagos Islands which are in the Gray Herbarium, as well as for advice and assistance in innumerable places, rendered doubly valuable on account of his intimate knowledge of the flora of these islands. Dr. Robinson has also been kind enough to read and criticise the manuscript and to give advice about the arrange- ment of the same. I wish also to express my thanks to Dr. W. G. Farlow of Harvard University for identifying the lichens and mosses, and to Miss Mary A. Day, Librarian of the Gray Herbarium, for assistance in looking up the rather large amount of literature made necessary in revising the nomenclature. I wish further to acknowledge the kindness of Prof. M. L. Fernald of the Gray Herbarium for assistance in many places, of Mr. Casimir de Candolle of Geneva, Switz- erland, for assistance on Peperomia, of Mr. A. S. Hitchcock of the U. S. Department of Agriculture for aid in regard to the Gramineae, of Mr. H. D. House for assistance in the identification of some of the members of the Convolvulaceae, of Mr. W. H. Ochsner, geologist of the expedition, and Mr. E. W. Gifford, joint ornithologist of the expedition, for infor- mation about their particular subjects, and of Mr. H. H. Bartlett of the U. S. Department of Agriculture for assistance in translating many of the descriptions of the new species, varieties, and forms into Latin. Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 11 ACCOUNT OF THE SPECIES OF VASCULAR PLANTS PTERIDOPHYTA FILICES Acrostichum L. A. aureum L. Sp. Pl. 1069 (1753); Rob. (1), 104.—ALBE- MARLE Isx.: Villamil, occasional in protected places at 3150 ft. (nos. 773-774). Further distr. general in tropical countries. Adiantum L. A. aethiopicum L. Sp. Pl. ed. 2, 1560 (1763); Rob. (1), 105.—Gatapacos Ips.: acc. to Moore. Further distr. general in tropical countries. A. Alarconianum Gaud. Voy. Bon. Bot. t. 99 (1846). A. incisum Presl, Rel. Haenk. I. 61, t. 10, £..3 (1830) ; Rob. (1), 105.—Garapacos Ibs.: acc. to Moore. Further distr. Mex., S. Am. A. concinnum H. & B. in Willd. Sp. V. 451 (1810) ; Rob. (1), 105.—Asinepon Ist.: common in lava cracks at 550 ft. (no. 776). ALBEMARLE IsL.: Cowley Bay, common among shady rocks at 2000 ft. (no. 779) ; Iguana Cove, common on side of the cliff above the cove (no. 780) ; Tagus Cove, com- mon in lava cracks at 1600 ft. (no. 778) ; Villamil, common in lava caverns at 1350 ft. (nos. 781-782). CHARLEs IsL.: on moist shady rocks at 1000 ft. (nos. 783-784). James IsL.: Darwin; Scouler. NarsoroucH Ist.: south side, Snodgrass and Heller. Further distr. Mex., W. Ind., northern S. Am. A. diaphanum BI. Enum. 215 (1828)—ALBeMarLeE ISL. : Villamil, occasional in moist places on the south side of the mountain at 3150 ft. (no. 785). Further distr. Old World. A. Henslovianum Hook. f. (3), 169; Rob. (1), 105.—As- INGDON Ist.: common in shady places 1500-1650 ft. (nos. 786-788). ALBEMARLE IsL.: Tagus Cove, occasional at 400 12 CALIFORNIA ACADEMY OF SCIENCES (Proc. 4ru Ser. ft., abundant at 4000 ft., (no. 790) ; Villamil, common in lava caverns at 1350 ft. (nos. 789, 791-793). CuHar es IsL.: Dar- win. CHATHAM Ist.: Wreck Bay, occasional at 1700 ft. (no. 794). INDEFATIGABLE IsL.: Academy Bay, common in shady places at 550 ft. (no. 795). James Ist.: James Bay, common on moist shady banks at 2150 ft. (nos. 796-797). Further distr. Andean S. Am. A.macrophyllum Sw. Prodr. 135 (1788).—ALBEMARLE Ist.: Villamil, common in lava caverns at 1350 ft. (no. 799). INDEFATIGABLE IsL.: Academy Bay, occasional in dense shade above 500 ft. (no. 800). James Ist.: James Bay, on shady banks at 2100 ft. (no. 801). Further distr. Mex., W. Ind., northern S. Am. This fern is always found in the densest shade where there is a considerable amount of moisture. A. parvulum Hook. f. (3), 168; Rob. (1), 106.—CHaRLEs Is_t.: Darwin. Endemic. A. patens Willd. Sp. V. 439 (1810); Rob. (1), 106.—Gat- APAGOS Ips.: acc. to Moore. Further distr. Mex., northern S. Am. A. petiolatum Desv. Berl. Mag. V. 326 (1811). A. Kaul- fussii Kunze, Linnaea, XXI. 221 (1848); Rob. (1), 105.— CuHaTHAM Isx.: acc. to Moore. INDEFATIGABLE IsL.: Acad- emy Bay, common in shady places at 500 ft. (no. 798). Fur- ther distr. Mex., W. Ind., S. Am. A. tetraphyllum H. B. Willd. Sp. V. 441 (1810). A. prio- nophyllum HBK. Noy. Gen. & Sp. I. 20 (1815); Rob. (1), 106.—CuHatuHam Isz.: acc. to Moore. Further distr. Mex. W. Ind., S. Am. Anogramma Link A. chaerophylla (Desv.) Link, Fil. Sp. 138 (1841). Gym- nogramme chaerophylla Desy. Berl. Mag. V. 305 (1811) ; Rob. (1), 109.—Cwartes Ist.: Darwin. Further distr. Mex., W. Ind., S. Am. Robinson, 1. c., expresses doubt as to the identity of the Darwin specimen. A. leptophylla (L.) Link, Fil. Sp. 137 (1841). Polypodium leptophylla L. Sp. Pl. 1092 (1753). Gymnogramme lepto- Vor. 1] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 13 phylla Desy. Jour. Bot. I. 26 (1813); Rob. (1), 109.— Cuartes Ist.: Baur. Widely distributed in tropical regions. Aspidium Sw. A. martinicense Spr. Anleit. III. 133 (1804). Nephrodiwm macrophyllum Bal. Syn. Fil. 300 (1874); Rob. (1), 110.— ALBEMARLE Ist.: Villamil, common in protected places on the south side of the mountain at 3150 ft. (no. 902). JAMEs Ist.: James Bay, common in moist situations at 2000 ft. (no. 901). Further distr. Mex., W. Ind., northern S. Am. Asplenium L. A. anisophyllum Var. latifolium, Hook. Sp. Fil. II. 111 (1860); Rob. (1), 106——Gatapacos Ins.: Capt. Wood. James Ist.: Darwin. Further distr. Mex., W. Ind., S. Am., Old World. A. cristatum Lam. Ency. II. 310 (1786). A. cicutarium Sw. Prod. 130 (1788); Rob. (1), 107.—Asrncpon Ist: common around 1950 ft. (no. 820). ALBEMARLE IsL.: Villa- mil, common in lava caverns at 1350 ft. (no. 821). CHaTHam Ist.: Wreck Bay, occasional in protected places around 1800 ft. (no. 822). INDEFATIGABLE IsL.: Academy Bay, common in shady places above 550 ft. (no. 823). JAmes IsL.: James Bay, rare in lava caverns at 1000 ft., common on moist shady rocks at 2150 ft.; (nos. 824-827). Further distr. Mex., W. Ind., S. Am., Africa. A. formosum Willd. Sp. V. 329 (1810) ; Rob. (1), 107.— AsIncpon Ist.: abundant at 1400 ft. (no. 825). ALBEMARLE Ist.: Iguana Cove, abundant in shade at 250 ft. (no. 833) ; Tagus Cove, common in lava crevices, 1600-2800 ft. (nos. 832, 834); Villamil, common among rocks at 1300 ft. (nos. 835-836). CuHartes Ist.: abundant on moist shady rocks at 1000 ft., and to some extent on the walls of the main crater at a somewhat higher elevation, (nos. 830, 844, 845). CHATHAM Iszt.: Wreck Bay, common in moist shady places at 650 ft. (no. 837). INDEFATIGABLE IsL.: Academy Bay, in leaf mold among rocks 400-600 ft., larger and more abundant at 14 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. the higher elevation, (nos. 841-842) ; northwest side, among rocks at 900 ft., (no. 1027) ; southeast side, common on shady rocks at 625 ft. (no. 843). JAmeEs IsL.: James Bay, abundant in lava caverns at 900 ft., and in moist shady places at 2150 ft., where it reaches a height of 18 inches, (no. 838). Nargor- ouGH IsL.: in the upper moist regions (no. 840). Further distr. general in tropical regions. A. laetum Sw. Syn. Fil. 79, 271 (1806) ; Rob. (1), 107.— CuatHaM Ist.: Capt. Wood. Further distr. Mex., W. Ind., S. Am. A. lunulatum Sw. Syn. Fil. 80 (1806); Rob. (1), 107.— CuHar es Isut.: Lee. Further distr. general in tropical regions. A. myriophyllum (Sw.) Presl, Rel. Haenk. I. 48 (1825). Caenopteris myriophylla Sw. Schrad. Jour. 1800, 2, 60 (1801). Asplenium rhizophyllum Kze. Linnaea, IX. 71 (1834) ; Rob. (1), 107.—Gatapacos Ips.: Capt. Wood. James Ist.: Dar- win. Further distr. general in tropical regions. A. praemorsum Sw. Prod. 130 (1788). 24 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. (no. 954). INDEFATIGABLE IsL.: Academy Bay, on trees, 375-450 ft., (no. 956); southeast side, rare on trees at 625 ft. James IsLt.: James Bay, on trees at 1300 ft. (no. 955). This fern is usually found in the transition and lower moist regions. Further distr. Mex., S. Am. P. loriceum L. Sp. Pl. 1086 (1753); Rob. (1), 112.— GaLapacos Ips.: Moore. Further distr. Mex., W. Ind., S. Am. P. pectinatum L. Sp. Pl. 1085 (1753); Rob. (1), 113.— ABINGDON IsL.: occasional among rocks in the wooded region above 1000 ft. (no. 961). ALBEMARLE IsL.: Cowley Bay, occurs first among rocks in shady woodland at 2000 ft. Below this elevation the soil is composed entirely of pumice, which is not well adapted to support a fern flora; Iguana Cove, among rocks in woodland near the shore; Tagus Cove, com- mon in lava caverns at 2200 ft. and on the west side of the mountain at 4000 ft. (no. 967); Villamil, common among rocks 100-3150 ft. (no. 969). CHARLES IsL.: common in lava crevices on the inner walls of the main crater at 1400 ft. (no. 968). CHatHAM Ist.: Wreck Bay, fairly abundant in shady woods at 700 ft. (no. 966). INDEFATIGABLE ISL.: Academy Bay, common in vegetable mold among rocks, 350- 500 ft., (no. 962); northwest side, occasional at 1000 ft.; southeast side, common among rocks at 625 ft. JAmeEs IsL.: James Bay, common above 1300 ft. (no. 966). This fern is most abundant in the lower part of the moist region but usually disappears when the vegetation becomes dense. Fur- ther distr. Mex., W. Ind., S. Am. P. percussum Cav. Prael. 243 (1801); Rob. (1), 113.— GaLapacos Ips.: Capt. Wood. Further distr. Mex., S. Am. P. Phyllitides L. Sp. Pl. 1083 (1753); Rob. (1), 113.— James Ist.: James Bay, common in open woodland above 1500 ft. (no. 969). “Large sword ferns” were reported from the upper regions of Aprncpon Ist., and Banks Bay, ALBE- MARLE Ist. by Mr. F. X. Williams, the entomologist of the expedition. From his description it seems very likely that it was this species that he saw. Further distr. S. U. S., Mex., W. Ind., S. Am. Vou.I1] STEWART—BOTANY OF THE GALAPAGOS ISLANDS PX5) P. pleiosoros Hook. f. (3), 166 (as pleiosorum) ; Rob. (1), 113.—James Ist.: Darwin. Endemic. P. polypodioides (L.) Hitchcock, Rep. Mo. Bot. Gard. IV. 156 (1893). Roba) lS — Aprncpon Ist.: common on rocks at 450 ft., occasional on the trunks and branches of trees on the upper parts of the island, (no. 978). ALBEMARLE IsL.: Iguana Cove, Snodgrass and Heller; Tagus Cove, common in lava crevices at 2800 ft. (no. 986) ; Villamil, common on recent lava near the shore and in woodland above 350 ft. BrnpLor IsL.: common on lava in the upper interior parts of the island (no. 971). CuHares Ist.: occasional in open. woodland at 1000 ft., abundant at 1400 ft., (no. 972). CHATHAM Ist.: Basso Point, first seen at 900 ft.; Wreck Bay, abundant in shady woodland at 600 ft., fairly common in open country at 1700 ft., occasional at 2100 ft., (nos. 973-974). DuNcAN Sw: occurs first at 1000 ft. on the north side of the island, and at 700 ft. on the south side where the vegetation is bathed by the fog-laden wind, common at 1300 ft. on the south side, (no. 975). Hoop Ist.: occasional on the southeast side of cliffs at 600 ft. (no. 979). INDEFATIGABLE Ist.: Academy Bay, occasional at 50 ft., abundant, covering rocks in open woodland, at 350-500 ft., (no. 985); northwest side, occa- sional at 650 ft.; southeast side, occasional at 600 ft., forming low brakes around 700 ft., (no. 987). James IsL.: James Bay, abundant on recent lava beds above 500 ft., apparently one of the first vascular plants to invade the recent lava above this elevation, (no. 977). Jervis Ist.: occasional in a very limited area around 1050 ft. (no. 984). NarsorouGH Ist. : south side, common in the upper regions. Further distr. Mex., W. Ind., northern S. Am. 26 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. P. thyssanolepis A. Br. Kl. Linn. XX. 392 (1847).— ALBEMARLE Ist.: Villamil, abundant at 600 ft., on trees at 1300 ft., (nos. 989-990). James Ist.: James Bay, occasional on the trunks of trees at 2750 ft. (no. 991). Further distr. S. U. S., Mex., W. Ind., northern S. Am. Polystichum Roth. P. aculeatum (L.) Schott, Gen. Fil. ad t. 9 (1834). Poly- podium aculeatum L. Sp. Pl. 1090 (1753). Aspidium acu- leatum Sw. Schrad. Jour. 1800, 2, 37 (1801)—ALBEMARLE Ist.: Villamil, occasional on the south side of the mountain at 3150 ft. (no. 802). Widely distributed. P. adiantiforme (Forst.) J. Sm. Hist. Fil. 220 (1875). Polypodium adiantiforme Forst. Prod. 82 (1786). Aspidium coriaceum Sw. Syn. Fil. 57 (1806); Rob. (1), 106.—ALBE- MARLE Ist.: Villamil, occasional on the south side of the mountain at 3150 ft. (no. 803). JAmes Ist.: Darwin. Fur- ther distr. W. Ind., S. Am., Old World. P. apiifolium (Sw.) C. Ch. Ind. 64 (1905); 578 (1906). Dicksonia apiifolia Sw. Schrad. Jour. 1800, 2, 91 (1801).— James Ist.: James Bay, occasional at 2000 ft. (nos. 882- 883). Further distr. Mex., W. Ind., Andean S. Am. Pteris L. P. aquilina var. esculenta Hook. f. Fl. N. Zeal. II. 25 (1855); Rob. (1), 114—Axrnepon Ist.: forms extensive brakes on the south side of the island above 1600 ft. (no. 992). ALBEMARLE IsL.: Iguana Cove, Snodgrass and Hel- ler; Villamil, common in open woodland, 1200-1400 ft.; also common on the southeast side of the mountain at 3150 ft., and on the floor of the crater at 2750 ft., where it forms exten- sive brakes, (no. 993). CHatuam Ist.: Wreck Bay, occurs first at 1000 ft., forming extensive brakes at 1700 ft., common at 2100 ft., (nos. 994-996). Further distr. general in tropical regions. P. propinqua var. Cumingiana Ag. Sp. Gen. Pterid. 65 (1839); Rob. (1), 115.—Gatrapacos Ips.: Capt. Wood. Further distr. Mex., northern S. Am. Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS Py Trachypteris André T. pinnata (Hook. f.) C. Chr. Ind. 634 (1906). Hemion- itis pinnata Hook. f. Trans. Linn. Soc. XX. 167 (1847). Acrostichum aureonitens Hook. f. Ic. Pl. X. t. 933 (1854) ; Rob. (1), 104—Astnepon Ist.: abundant on rocks at 600- 1000 ft. (no. 772). Arpemarte Ist.: Cowley Bay, common around 2000 ft.; Iguana Cove, abundant on shady rocks near the shore (no. 768); Tagus Cove, abundant at 500-4000 ft. (nos. 766-767) ; Villamil, on rocks in shady places, 100-1300 ft.;) (no: 771). CHartEs Ist.: common on moist rocks at 1000 ft. (no. 1026). CHatHam Ist.: Basso Point, on rocks at 900 ft.; Wreck Bay, occasional at 900 ft. INDEFATIGABLE Ist.: Academy Bay, occasional on the sides of steep bluffs at 50 ft., abundant above 350 ft., (no. 766); southeast side, rare at 450 ft., abundant above 500 ft. JAmes Ist.: James Bay, on rocks in open woodland, 800-1300 ft., (no. 770). NarporouGH Is.: south side, Snodgrass and Heller. Trichomanes L. T. pusillum Sw. Prod. 136 (1788).—James Ist.: James Bay, common on moist tufa walls at 2050 ft. (no. 1012). Further distr. Mex., W. Ind., northern S. Am., Africa. Vittaria J. Sm. V. angustifolia (Sw.) Bak. Fl. Bras. I. 2, 544 (1870). Pteris angustifolia Sw. Prod. 129 (1788). Taenitis angusti- folia R. Br. Prod. 154, in note (S10); Rob. (1), 115. Gatapacos Ips.: Capt. Wood. Further distr. Mex., W. Ind., S. Am. SALVINIACEAE Azolla Lam. A. caroliniana Willd. Sp. V. 541 (1810); Rob. (1),.115.— Cuartes Ist.: abundant on mud and floating in water around springs at 1000 ft. (no. 3441). CHatuam Ist.: Wreck Bay, common in small streams, 1000-1700 ft., (no. 3442). Further distr. U. S., Mex., S. Am. 28 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. Salvinia L. S. sp. Wolf, (1), 284; Rob. (1), 115.—Cuartes Ist.: in brooks near the hacienda, acc. to Wolf 1. c. EQUISETACEAE Equisetum L. E. bogotense HBK. Nov. Gen. & Sp. I. 42 (1815).—ALBeE- MARLE Ist.: Villamil, rare on the south rim of the crater at 3150 ft. (no. 3443). Further distr. Mex. (Cent. Am.), W. Ind., S. Am. LYCOPODIACEAE Lycopodium L. L. clavatum LL; Sp: Pl. 1101 (1753); Reb: (1); 115— CuHatHamM Ist.: Wreck Bay, forming thick tangled masses 2-3 ft. high above 1500 ft. (no. 1014). James Ist.: James Bay, rare around 2500 ft. (no. 1015). Widely distributed. L. complanatum L. Sp. Pl. 1104 (1753).—ALBEMARLE Ist.: Villamil, occasional at 3150 ft. (no. 1016). Widely distributed. L. dichotomum Jacq. Enum. Vindob. 314 (1762); Rob. (1), 115.—Avpemarte Ist.: Villamil, common on the higher branches of trees, 500-700 ft., (no. 1017). INDEFATIGABLE Ist.: Academy Bay, on the trunks and branches of trees, 400- 500 ft., (no. 1019); northwest side, common on the higher branches of Psidiwm galapageium trees above 1000 ft. Fur- ther distr. Mex., W. Ind., S. Am., Madagascar. L. reflexum Lam. Encye. III. 653 (1789).—ALBEMARLE Ist.: Villamil, common at 3150 ft. (no. 1020). Further distr. Mex., W. Ind., S. Am. L. taxifolium Sw. Fl. Ind. Occ. II. 1573 (1806).— Cuares Ist.: abundant on branches of Acnistus ellipticus trees at 1700 ft. (no. 1021). Wolf 1. c. 283 refers to two un- determined species of Lycopodium from this island and it seems likely that this is one of them. JAmes Ist.: James Bay, Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 29 common on the branches of Zanthoxylum Fagara trees at 2150 ft. (nos. 1023-1024). Further distr. Mex., W. Ind., northern S. Am. SPERMATOPHYTA MONOCOTYLEDONEAE POTAMOGETONACEAE Potamogeton Tourn. Ps pectinatus L. Sp. Pl. 127 (1753) 3 Rob. (1), 115. —=Ar- BEMARLE Isi.: Iguana Cove, Snodgrass and Heller. Widely distributed. Ruppia L. R. maritima L. Sp. Pl. 127 (1753) ; Rob. (1), 116.—ALBE- MARLE Isi.: Elizabeth Bay, Snodgrass and Heller. Widely distributed. NAJADACEAE Najas L. N. marina L., var. latifolia A. Br. ex Schum. in Mart. FI. Bras. III. pt. 3, 725 (1894); Rob. (1), 116—ALBEMARLE Ist.: Iguana Cove, Snodgrass and Heller. Further distr. tropical S. Am. GRAMINEAE Ammophila Host. A, arenaria (L.) Link, Hort. Berol. I. 105 (1827). Arundo arenaria L. Sp. Pl. 82 (1753).—Srymour Ist., sourH: form- ing a patch about one-fourth mile long on a sand beach on the west side of the island. The specimen is sterile and somewhat doubtful as to species, (no. 1195). Further distr. N. Am., Europe. Anthephora Schreb. A. hermaphrodita (L.) O. Kze. Rev. Gen. II. 759 (1891). Tripsacum hermaphroditum L. Sp. Pl. ed. 2, 1397 (1763). An- thephora elegans Schreb. Beschr. Gras. II. 105 t. 44 (1810); Rob. (1), 116.—A.bBemarte Ist.: Tagus Cove, occasional in 30 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. tufaceous soil in open sunny places on the lower parts (no. 1196). Cartes IsL.: rare at 850 ft., common in tufaceous soil at 1200 ft., (nos. 1197-1198). CHatHam Ist.: north side, Baur. INDEFATIGABLE IsL.: northwest side, Andersson. James Isi.: James Bay, Snodgrass and Heller. Further distr. Mex., W. Ind., S. Am. Aristida L. A. divulsa Anderss. (1), 143, and (2), 49; Rob. (1), 116.— Axstncpon Ist.: Snodgrass and Heller. BinpiLoe Ist.: Baur. CHATHAM Ist.: Sappho Cove, occasional in bunches on the recent lava in the vicinity of the cove, (no. 1200). James IsL.: James Bay, common in lava crevices near the shore (no. 1199). Endemic. A. repens Trin. Mém. Acad. Pétersb. ser. VI. I. 87 (1831); Rob. (1), 117.—James Ist.: Douglas. Endemic. A. subspicata Trin. & Rupr. Mém. Acad. Pétersb. ser. VI. I. 125 (1842); Rob. (1), 117.—Asrinepon Ist.: common in lava crevices in the vicinity of the shore and also sparingly at 1100 ft. (nos. 1201-1203). ALBEMARLE IsL.: Cowley Bay, common in soil of pumiceous origin at 1000 ft. (no. 1207) ; Elizabeth Bay, Snodgrass and Heller; Iguana Cove, common on the side of the cliffs above the cove, also common in open places around 300 ft., (no. 1206); Tagus Cove, common in open country near the shore and around 1000 ft. (no. 1208) ; Villamil; common in lava crevices near the shore (nos. 1204- 1205). Barrincton Ist.: occasional in loose soil among masses of lava (no. 1210). Brnpior IsL.: common near the shore and on the crests of tufa ridges in the interior of the island (no. 1211). Brattie Ist.: (no. 1212). CHARLES Ist.: abundant at 850 ft., occasional at 1750 ft., (nos. 1213, 1214, 1216, 1217); Cormorant Bay, occasional in lava crev- ices (no. 1215). CHatTHAm Ist.: Basso Point, occasional on lava and in open places in the dry region where the soil is very loose in texture (no. 1218). Garpner Ist. (near Hoop Ist.) : common everywhere (no. 1219). Hoop Ist. : occasional among rocks (no. 1220). INDEFATIGABLE IsL.: Academy Bay, abundant in open places in the vegetation on the lower parts (no. 1221); north side, abundant in ashy soil on the Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 31 lower parts (no. 1224); northeast side, abundant on the flat area near the shore (no. 1222); northwest side, common in tufaceous soil (no. 1225). James Ist.: James Bay, Snodgrass and Heller; northeast side, occasional in sand and in lava crev- ices (no. 1226). NarsporouGH Ist.: north side, abundant in lava crevices (no. 1227). Srymour Ins., NorTH and souUTH: Snodgrass and Heller. Further distr. S. Am. This is one of the most abundant and wide spread grasses of the dry region. It occurs commonly where the soil is too porous to support much large vegetation, and in such situations it often covers considerable areas. A. villosa Rob. & Greenm. (1), 144, 149; Rob. (1), 117.— Duncan Ist.: abundant on the lower and dry parts of the island (nos. 1228-1229). Jervis Ist.: Baur. Endemic. Bouteloua Lag. B. pilosa (Hook. f.) Benth. acc. to Watson, Proc. Am. Acad. XVIII. 179 (1883). Eutriana pilosa Hook. f. (3), 173. B. pilosa Benth. |. c.: Rob. (1), 117—Asrtnepon Ist.: occa- sional in open places at 1050 ft. (no. 1230). ALBEMARLE Ist. : Iguana Cove, in spreading bunches among thick vegetation at 250 ft. (no. 1231) ; Tagus Cove, abundant in tufaceous soil on the lower parts of the island (no. 1232); Villamil, occasional in woodland at 250 ft. (no. 1233). Barrrneton Isi.: Snod- grass and Heller, CuatHam Ist.: north side, Andersson. INDEFATIGABLE IsL.: north side, Snodgrass and Heller. James Ist.: James Bay, Snodgrass and Heller. Jervis Isu.: Baur. Narsoroucu Ist.: north side, abundant on lava beds near the shore (no. 1234). Srymour Ips., NorTH and souTH: Snodgrass and Heller. Endemic. Cenchrus L. C. distichophyllus Griesb. Cat. Pl. Cub. 234 (1866); Rob. (1), 118.—ALBeEmart-e IsL.: Villamil, Baur. Cuatuam Ist.: Sappho Cove, forming dense mats on sand beaches (no. 1236). Hoop Ist.: fairly common on sand beaches (no. 1235). Fur- ther distr. Cuba. C. granularis Anderss. (1), 140, and (2), 47; Rob. (1), 118.—ALBEMARLE IsL.: Tagus Cove, common at 100-4000 ft. 32 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. (nos. 1237-1238). CHAR Es IsL.: rare at 900 ft. (no. 1240). CuaTHAM Ist.: Andersson; A. Agassiz. NaArporouGH IsL.: north side, abundant on lava beds (no. 1239). Srymour Ist., soutH: Snodgrass and Heller. C. platyacanthus Anderss. (1), 139, (2), 47; Rob. (1), 118. —Asincpon Ist.: occasional on the lower parts (no. 1241). ALBEMARLE Ist.: Iguana Cove, in spreading bunches on the sides of the cliffs above the cove (no. 1242). BArrINGTON Ist.: Snodgrass and Heller. Brtnvtor Ist.: Snodgrass and Heller. Brattre Ist.: (no. 1244). Cares Isu.: Snod- grass and Heller. CHATHAM IsL.: Wreck Bay, abundant near the shore and in open places in the vegetation to 150 ft. (nos. 1245-1246). GarpNeR Ist. (near Hoop Ist.): Snodgrass and Heller. Woop Ist.: Baur. INDEFATIGABLE IsL.: Acad- emy Bay, abundant in sandy soil near the shore; north side, Snodgrass and Heller. JAMeEs Ist.: northeast side, occasional in lava crevices; JAMES Bay, Snodgrass and Heller. NArBor- ouGH IsL.: Snodgrass and Heller. Endemic. C. sp. Rob. (1), 118.—Cuatuam Ist.: Snodgrass and Hel- ler. Chloris Sw. C. anisopoda Rob. (1), 118.—Cuartes IsL.: occasional among rocks near the shore (no. 1250). INDEFATIGABLE ISL. : north side, fairly abundant in lava crevices near the shore (no. 1251); southeast side, on the lower dry parts (no. 1252). Endemic. C. elegans HBK. Nov. Gen. & Sp. I. 166, t. 49 (1815); Rob. (1), 119.—Cuartes Ist.: Snodgrass and Heller. Sry- MOUR IsL., NorTH: Snodgrass and Heller. Further distr. S. Wr UEtss" Mex C. radiata (L.) Sw. Prodr. 26 (1788). Agrostis radiata L. Syst. 10, ed. 2, 873 (1759). C. radiata Sw. 1. c.; Rob. (1), 119.—Cuartes Ist.: Baur. Further distr. W. Ind., S. Am. Dactyloctenium Willd. D. aegyptium (L.) Richter, Plan. Eur. I. 68 (1890). Cyno- surus aegyptius L. Sp. Pl. 72 (1753). Eleusine aegyptica Desf. Fl. Alt. I. 85 (1798); Rob. (1), 119.—Cuartes Ist. : Vor. 1 STEWART—BOTANY OF THE GALAPAGOS ISLANDS 33 abundant near the shore (no, 1253). CHatTHam Ist.: Wreck Bay, abundant in dry sandy soil near the beach (no. 1254). Hoop Ist.: Baur. INDEFATIGABLE IsL.: Academy Bay, abundant in sand near the shore (no. 1255). Widely distrib- uted. Digitaria Scop. D. sanguinalis (L.) Scop. F. Carn. ed. 2, 1, S2EAA7Z2 Panicum sanguinale L. Sp. Pl. 57 (1753); Rob. (1), 123.— CuHartes Ist.: occasional in open places at 1100 ft. (no. 1304). CuatrHam Ist.: Wreck Bay, common in open sunny places to 450 ft. (nos. 1303, 1305). Of wide distribution. Eleusine Gaertn. E. indica Gaertn. Fruct I. 8 (1788); Rob. (1), 120.— Cuartes Ist.: abundant in open meadows-at 1000 ft., also common at 1500-1750 ft., (nos. 1255-1258). CHATHAM IsL.: Wreck Bay, abundant at 300-800 ft., occasional at 1300 ft., (nos. 1259-1261). Widely distributed. Probably an intro- duced species on the islands. Eragrostis Host. E. bahiensis Roem. & Sch. Mant. II. 318 (1824); Rob. (1), 120.—Asinopon Isx.: occasional in bunches around 1000 ft. (no. 1262). ALBEMARLE Ist.: Iguana Cove, Snodgrass and Heller. Further distr. S. Am. E. ciliaris (L.) Link, Hort. Berol. I. 192 (1827). Poa cili- aris L. Syst. ed. 10, 875 (1760). Eragrostis ciliaris Link, 1. c. ; Rob. (1), 120—Asrnepon Ist.: Snodgrass and Heller. At- BEMARLE Ist.: Macrae; Tagus Cove, Snodgrass and Heller. Brnpcor Ist.: on the upper parts (no. 1263). CHARLEs ISL. : abundant 430-1100 ft. (nos. 1264-1265). CHATHAM IsL.: Darwin; Andersson; Snodgrass and Heller. Hoop Ist.: Snodgrass and Heller. James Ist.: James Bay, common in dry places on sides of cliffs around the bay (no. 1266). Nar- BOROUGH IsL.: north side, abundant on recent lava beds (no. 1267). Tower Isu.: Baur; Snodgrass and Heller. Widely distributed. E. megastachya (Koehl.) Link, Hort. Berol. I. 187 (1827). Poa megastachya Koehl. Descr. Gram. 181 (1802). Eragros- 34 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. tis major Host. Gram. IV. 14, t. 24 (1809) ; Rob. (1), 120.— BARRINGTON Isi.: Snodgrass and Heller. BRAtTTLE IsL.: (no. 1268). CuHarR LES Is~t.: common above 450 ft. (nos. 1269- 1270). CHatTHam Ist.: Wreck Bay, common in dry sandy soil near the shore (no. 1271). Duncan Ist.: occasional near the shore. GARDNER IsL. (near Hoop Ist.): Snodgrass and Heller. Hoop Isut.: Snodgrass and Heller. INDEFATIGA- BLE Ist.: Academy Bay, fairly abundant near the shore (no. 1272). Seymour Ist., sourH: Snodgrass and Heller. Tower Ist.: Snodgrass and Heller. WrENMAN IsL.: (no. 1273). Widely distributed. E. pilosa(L.) Beauv. Agrost. 71 (1812). Poa pilosa L. Sp. Pl. 68 (1753). E. pilosa Beauv. 1. c.; Rob. (1), 120—James Ist.: Darwin. Widely distributed. Eriochloa HBK. E. distachya HBK. Nov. Gen. & Sp. I. 95, t. 30 (1815); Rob. (1), 121—Cuatuam Ist.: Snodgrass and Heller. Fur- ther distr. northern S. Am. E. punctata (L.) Desv. in Ham. Prod. Pl. Ind. Occ. 5 (1825). Milinm punctatum L. Sp. Pl. ed. 2, 91 (1763).— CuatHAM Ist.: Basso Point, occasional in open places at 900 ft. (no. 1274). Further distr. U. S., W. Ind., S. Am. Leptochloa Beauv. L. albemarlensis Rob. & Greenm. (1), 145, 149; Rob. (1), 121.—Arinepon IsL.: common on lava beds at 450 ft. (no. 1275). ALBEMARLE IsL.: Villamil, Baur; Iguana Cove, Snod- grass and Heller. Endemic. L. filiformis (Lam.) Roem. & Sch. Syst. II. 580 (1817). Festuca filiformis Lam. Illust. I. 191 (1791). L. filiformis Roem & Sch. 1. c.; Rob. (1), 121.—Duncan Isv.: Snodgrass and Heller. SryMourR Ist., soutH: Snodgrass and Heller. Widely distributed in tropical regions. L. Lindleyana Kunth, Rev. Gram. II. 655, t. 215 (1829) ; Rob. (1), 121.—Asrinepon IsL?: occasional on lava beds in the lower dry region (no. 1276). ALBEMARLE IsL.: Macrae; Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 35 Cowley Bay, Baur; Tagus Cove, (no. 1277). BinpLoe Ist.: Snodgrass and Heller. CuatHam Ist.: Andersson. Nar- BOROUGH IsL.: north side, abundant on lava beds (no. 1278). Endemic. L. mucronata (Michx.) Kunth, Rev. Gram. I. 91 (1829). Eleusine mucronata Michx. Fl. Bor. Am. 65 (1803). L. mu- cronata Kunth, |. c.; Rob. (1), 121.—BarrincTon Ist.: (no. 1279). Garpner Ist., (near Hoop Isu.): (no. 1281). Hoop IsL.: common in scant soil among rocks (no. 1280). Further distr. S. U. S., Mex., W. Ind., S. Am. L. virgata (L.) Beauv. Agrost. 71 (1812). Cynosurus vir- gatus L. Sp. Pl. ed. 2, 106 (1762). L. virgata Beauv. 1. c.; Rob. (1), 121.—Cuartes Ist.: upper grassy region acc. to Andersson. CHATHAM Ist.: Wreck Bay, occurs first at 650 ft. (no. 1282). Further distr. Mex., W. Ind., S. Am. Oplismenus Beauv. O. setarius (Lam.) Roem. & Sch. Syst. II. 481 (1817). Panicum setarium Lam, Il. 1. 170 (1791). O. setarius Roem. & Sch. 1. c.; Rob. (1), 121—Cuatuam Ist.: Wreck Bay, in shady places in cultivated ground around 1000 ft. (no. 1283). Probably an introduced species. Further distr. U. S., Mex., W. Ind., S. Am. Panicum L. P. colonum L. Syst. ed. 10, 870 (1760); Rob. (1), 122.— Cuartes Ist.: Darwin; Andersson. Further distr. general in tropical regions. P. fasciculatum Sw. Prodr. 22 (1788) ; Rob. (1), 122. P. fuscum Sw. Prodr. 23 (1788); Rob. (1), 122.—ALBEMARLE Ist.: Turtle Cove, in dense patches 4-6 ft. high in moist places near the shore (no. 1284); Villamil, in dense patches 5-6 ft. high in low places 2-3 miles back from the beach. The soil in these areas is kept moist the greater part of the time by the underflow of water from the interior of the island, (nos. 1285- 1286). CHarues Ist.: Snodgrass and Heller. CHATHAM Ist.: Wreck Bay, abundant on the sides of the road leading to the hacienda. This grass was only seen in January and Feb- ruary, when there is considerable water standing in the low 36 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. places where it occurs, (no. 1286). INDEFATIGABLE IsL.: north side, Andersson. JAmeEs Ist.: James Bay, Snodgrass and Heller. Further distr. Mex., W. Ind., S. Am. P. geminatum Forsk. Fl. Aeg.-Arab. 18 (1775). P. fluitans Retz. Obs. HI. 8 (1783); Rob. (1), 122.—Cuartes Is.: abundant on the sides of a moist cliff above a spring at 1000 ft. (nos. 1287-1288). CHatHAmM Ist.: Wreck Bay, common in large bunches at 450 ft. (no. 1290). Duncan Ist.: common in bunches at 900 ft. (no. 1291). Hoop Isy.: on the margin of a mud lake ace. to Snodgrass and Heller. The dry culms of a grass, probably this one, were noticed in the dry bed of this lake in June. Further distr. W. Ind., S. Am., Old World. P. hirticaulum J. & C. Presl, Rel. Haenk. I. 308 (1830) ; Rob. (1), 122.—Barrineton Ist.: Snodgrass and Heller. Cuartes Isu.: Andersson. CHATHAM IsL.: Wreck Bay, common in open places on the lower parts of the island, (no. 1293). GARDNER IsL., (near Hoop Isx.): dried remains of this grass were found in June (no. 1294). Hoop Ist.: abund- ant (no. 1297). INDEFATIGABLE IsL.: Academy Bay, abund- ant (no. 1297); north side, Snodgrass and Heller; northeast side, common on the table land above the beach (no. 1296). SEYMouR Ips., NorRTH and souTH: Snodgrass and Heller. Robinson, |. c., suggests that this grass is a recent introduction to the islands. It might be mentioned in this connection that the islands on which it occurs are frequently visited by the inhabitants of both Albemarle and Chatham Ids. Further distr. Mex. Var. minus Anderss. (1), 135, and (2), 44; Rob. (1), 123. —Cuar Es Isu.: Andersson; Snodgrass and Heller. CHaAtTH- AM Ist.: Andersson. Endemic. P. molle Sw. Prod. 22 (1788); Rob. (1), 123—CHaTHAM Ist.: Chierchia. Further distr. W. Ind., S. Am., tropical Asia. P.-multiculmum Anderss. (1), 133, and (2), 43; Rob. (1), 123.—ALBEMARLE IsL.: Iguana Cove, Snodgrass and Heller; Turtle Cove, occasional in scant soil near the shore (no. 1298) ; Villamil, occasional in lava crevices near the shore (no. 1299). BarrInGcTON Ist.: Snodgrass and Heller. CHARLES IsL.: in open places near the shore (no. 1300). CHATHAM IsL.: Vor. I) STEWART—BOTANY OF THE GALAPAGOS ISLANDS 37 Wreck Bay, occasional in low places around 300 ft. The soil is very wet in this region during the season in which the species occurs, (no. 1301). Duncan Ist.: Snodgrass and Heller. GARDNER IsLt. (near Hoop Ist.): Snodgrass and Heller. Hoop Isi.: common among rocks (no. 1302). Endemic. P. serotinum (Michx.) Trin. Gram. Panic. 166 (1826). Digitaria serotina Michx. Fl. I. 46 (1803). P. serotinum Trin. 1. c.; Rob. (1), 123.—Cuar tes Ist.: Edmonston. Rob- inson |. c. suggests that there may have been some mistake in the identification of this specimen. Further distr. S. U.S. P. sp.—Cuar es Ist.: (no. 1306). P. sp.—CuatTuam Ist.: (no. 1308). P. sp—Duncan Ist.: (no. 1307). All three of the above specimens are in too poor a condition for determination. They probably represent three distinct species. Paspalum L. P. canescens Anderss. (1), 132, and (2), 42; Rob. (1), 123. P. longe-pedunculatum Rob. (1), 124, not Le Conte-—ALBE- MARLE Ist.: Cowley Bay, Andersson; Iguana Cove, in bunches on the sides of the cliffs above the cove (nos. 1317-1318) ; Tagus Cove, in large bunches, 600-4000 ft., (nos. 1309, 1319- 1320); Villamil, common at 3150 ft. (no. 1311). BrnpdLoe Ist.: Snodgrass and Heller. CHaARtEs Ist.: Darwin; Anders- son. CHATHAM Ist.: Wreck Bay, occasional, 450-2100 ft., (nos. 1321-1322). NarsBorouGu Ist.: north side (no. 1310). Endemic. P. conjugatum Berg. Act. Helv. VII. 129, t. 8 (1772) ; Rob. (1), 123—ALBeMar_e Ist.: Villamil, one of the commonest grasses in the region above 1500 ft. It also occurs to some extent below this elevation, (no. 1312). CHARLEs IsL.: com- mon on the sides of moist tufa walls around 1000 ft. (no. 1312). CuHatHam Ist.: Wreck Bay, common in the grassy country above 700 ft. (nos. 1314-1315). James IsL.: James Bay, abundant in open places and in open woodland above 1500 ft. (no. 1316). This is the principal forage grass on the islands where it occurs. Further distr. Mex., W. Ind., S. Am., Tropical Africa. 38 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. P. distichum L. Amoen. Acad. V. 391 (1760); Rob. (1), 123.—JameEs Ist.: Orchilla Bay, Bawr. Further distr. tropical and subtropical regions. P. penicillatum Hook. f. (3), 171; Rob. (1), 124.— Cuar-eEs Isxt.: Darwin. Endemic. P. scrobiculatum L. Mant. I. 29 (1767); Rob. (1), 124.— CuHatHam Ist.: Chierchia. Further distr. general in tropics of old world. P. sp. Rob. (1), 124.—InpErFaTiIcABLeE IsL.: north side, Snodgrass and Heller. Probably a new species, acc. to Robin- son, line: Pennisetum Rich. P. exalatum (Anderss.) Hook. f. & Jacks. Ind. Kew. I. 112 (1893). Amphochaeta exalata Anderss. (1), 137, (2), AD it. Ie f.2. P. pauperum Nees acc. to Steud. Syn. 102 (1855) ; Rob. (1), 119.—Avsemarte IsL.: Cowley Bay, Andersson; Iguana Cove, occasional in dense patches 6-7 ft. high on the sides of the cliffs above the cove (no. 1323); Tagus Cove, a considerable area is covered with a dense growth of this grass around the top of the mountain, 3850-4000 ft., (no. 1324). The dried culms, collected at Elizabeth Bay and ascribed to Chusquea sp. by Robinson, (1), 119, no doubt belong to this species. Nar- BOROUGH Ist.: Mr. R. H. Beck reported a heavy growth of grass around the top of this island. From his description it is probably of this species. Endemic. Setaria Beauv. S. floriana Anderss. (1), 138, (2), 46; Rob. (1), 124.— CHARLES Isut.: Andersson. Endemic. S. setosa (Sw.) Beauv. Agrost. 51 (1812). Panicum seto- sum Sw. Prodr. 22 (1788). S. setosa Beauv. 1. c.; Rob. (1), 124.—ALBEMARLE IsL.: Elizabeth Bay, Snodgrass and Heller ; Iguana Cove, (no. 1091) ; Tagus Cove, abundant in tufaceous soil around the base of the mountain (no. 1090); Villamil, occasional in bunches on the lower parts (no. 1092). Bar- RINGTON Isz.: Snodgrass and Heller. CHarvers Ist.: abund- ant around 1750 ft. (nos. 1293-1294). CuatTHam Ist.: Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 39 Sappho Cove, on lava beds near the shore (no. 1295) ; Wreck Bay, grows in shady places at 250-600 ft. (nos. 1296-1297). Duncan Ist.: dried remains (no. 1179). Hoop Isu.: occa- sional around 600 ft. (nos. 1180-1181). InprraticAste Ist. : Academy Bay, abundant in open woodland around 350 ft.; southeast side, in rather open country around 500 ft. (no. 1182). James Ist.: James Bay, common on the lower parts (no. 1183). Narsoroucu Ist.: Snodgrass and Heller. Fur- ther distr. tropical regions. S. n.sp-? Hook. f. (3), 172; Rob. (1), 125.—ALBEMARLE Isu.: Macrae. Endemic. Sporobolus R. Br. S. domingensis (Trin.) Kunth, Enum. I. 214 (1833). Vilfa domungensis Trin. in Spreng. neue Ent. II. 59 (1821). S. domingensis Kunth 1. c.; Rob. (1), 125—Asinepon Ist: Snodgrass and Heller. ALBEMARLE Ist.: Iguana Cove, Snod- grass and Heller; Tagus Cove, occasional on the lower parts and at 4000 ft. (nos.-1184-1185). Hoop Isu.: Snodgrass and Heller. Further distr. Mex., W. Ind. S. indicus (L.) R. Br. Prod. I. 170 (1810). Agrostis indica es Spy ls 63'(1753)): "Ss andicusse Berle: Rop: ()).125—— ALBEMARLE Ist.: Villamil, common at 3150 ft. (no. 1186). CHARLES Isx.: occasional at 1600 ft. (no. 1187). CHatHam Ist.: Wreck Bay, occasional in bunches at 1700 ft. (no. 1188). Widely distributed. S. virginicus (L.) Kunth, Rev. Gram. I. 67 (1829). Agros- tis virginica L. Sp. Pl. 63 (1753). S. virginicus Kunth, 1. c.; Rob. (1), 125.—A.sBemarte Ist.: Elizabeth Bay, Snodgrass and Heller; Villamil, covers a considerable area on the low flat just back of the beach (no. 1189). CHartes Ist.: common on sand beaches (no. 1190). CuHatHam Ist.: Andersson; A. Agassiz; Snodgrass and Heller. INDEFATIGABLE Ist.: Acad- emy Bay, in dense mats on sand beaches (no. 1191) ; southeast side, common on the shore and around the borders of salt water lagoons where the soil is strongly impregnated with salt (no. 1191). James Ist.: northeast side, on sand beaches. Further distr. U. S., Mex., W. Ind., S. Am. January 11, 1911 40 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. Stenotaphrum Trin. S. secundatum ( Walt.) O. Kze. Rev. Gen. II. 794 (1891). Ischaemum secundatum Walt. Fl. Car. 249 (1788). Stenota- phrum glabrum Trin. Fund. Agrost. 176 (1820); Rob. (1), 126.—ALBEMARLE IsL.: Iguana Cove, Snodgrass and Heller. CHATHAM Ist.: Wreck Bay, forms thick mats in open country around 650 ft. (no. 1193). Further distr. tropical shores of both continents. Stipa L. S. rostrata Anderss. (1), 142, (2), 48; Rob. (1), 126.— CHARLES IsL.: in open places near the beach, evidently a younger specimen than the one described by Andersson, (no. 1194). CHatTHAmM IsL.: Andersson. Endemic. CYPERACEAE Cyperus L. C. aristatus Rottb. Descr. Nov. Pl. 23, t. 6, f. 1 (1786) ; Rob. (1), 126—ALBEMARLE IsL.: Iguana Cove, common in lava cracks near the shore (no. 1028); Tagus Cove, common on lava beds at 4000 ft. (no. 1027). CuHartes Ist.: Darwin. CHATHAM Is.: Wreck Bay, in swampy places, 1000-1750 ft., (no. 1029). James Ist.: Scouler. NArsBorouGH Isi.: Snod- grass and Heller. Widely distributed. C. brachystachys Anderss. (2), 53, t. 13, f. (2); Rob. (1), 126.—AnincGpoN IsL.: occasional at 600 ft., common in open places in the vegetation at 1400 ft., (no. 1030). ALBEMARLE IsL.: Iguana Cove, Snodgrass and Heller; Tagus Cove, com- mon above 300 ft. (no. 1032) ; Villamil, Baur. CHartes IsL. : occasional on rocks near the shore (no. 1033). CHATHAM Ist.: Baur. Duncan IsLt.: common around 1300 ft. (no. 1034). James IsL.: occasional on lava at 850 ft. (no. 1035). Jervis Ist.: (no. 1036). Tower Ist.: Snodgrass and Heller. Endemic. C.. confertus Sw. Prodr. 20 (1788) ; Rob. (1), 127.—ALBE- MARLE IsL.: Iguana Cove, common to above 400 ft.; Tagus Cove, occasional on lava beds, 100-600 ft., (nos. 1038-1039) ; Villamil, occasional at 650 ft. (no. 1040). Brinpxoe Ist. : Snodgrass and Heller. CHARLES IsL.: occasional at 1000- Vor. IJ STEWART—BOTANY OF THE GALAPAGOS ISLANDS 41 1250 ft. (nos. 1042-1244). CHatrHam Ist.: Wreck Bay, on rocks at 500 ft. (no. 1041). Duncan Isi.: Snodgrass and Heller. Hoop Ist.: Snodgrass and Heller. UNDEFATIGABLE Ist.: northwest side, Andersson. JAMES IsL.: Andersson. Widely distributed. C. esculentus L. Sp. Pl. 45 (1753) ; Rob. (1), 127.—ALBE- MARLE Ist.: Tagus Cove, occasional on rocky cliffs at 100 ft. (no. 1045). CHatHam Ist.: Wreck Bay, occasional in sandy soil near the shore (no. 1046). Widely distributed. C. galapagensis Caruel (1), 621; Rob. (1), 127—Cuatu- AM IsL.: Chierchia. Endemic. C. grandifolius Anderss. (1), 157, (2), 56; Rob. (1), 127. —CHARLES IsL.: occasional among rocks at 1550 ft. (no. 1047). CHatHam Ist.: Wreck Bay, common in large bunches above 1800 ft. (no. 1048). Endemic. C. laevigatus L. Mant. II. 179 (1771); Rob. (1), 127.— ALBEMARLE IsL.: Elizabeth Bay, Snodgrass and Heller; Villa- mil, occasional in brackish pools on the lower parts of the island (no. 1049). INDEFATIGABLE IsL.: southeast side, abundant in brackish pools near the shore (no. 1050). Further distr. tropical regions. C. ligularis L. Amoen. Acad. V. 391 (1760) ; Rob. (1), 127. —ALBEMARLE Iszt.: Elizabeth Bay, Snodgrass and Heller; Iguana Cove, Snodgrass and Heller; Tagus Cove, Snodgrass and Heller; Villamil, common in lava crevices near the shore and to some extent around 650 ft. (nos. 1051-1053). BinDLoE Isut.: (no. 1054). INDEFATIGABLE IsL.: Academy Bay, in pools of slightly brackish water near the shore (no. 1055). Further distr. W. Ind., tropical S. Am., Africa. C. Mutisii (HBK.) Anderss. (2), 53. Mariscus Mutisii HBK. Nov. Gen. & Sp. I. 216, t. 66 (1815). Cyperus Mutisii Anderss. 1. c.; Rob. (1), 128.—Asinepon Ist.: abundant around steam jets at 1000 ft. (no. 1056). ALBeMarte IsL.: Elizabeth Bay, Snodgrass and Heller; Iguana Cove, abundant in open places in the vegetation above the cove (no. 1057) ; Tagus Cove, (nos. 1059-1061) ; Villamil, common in lava crev- ices on the lower parts and at 500 ft. BINDLOoE IsL.: occasional 42 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. in tufaceous soil on the lower parts (no. 1062). Cartes Ist. : rare among rocks at 50 ft., common at 1450 ft., (no. 1063). CHATHAM Ist.: Andersson; Baur. GARDNER Ist. (near Hoop Ist.): (no. 1065). INDEFATIGABLE IsL.: Academy Bay, occasional at 100 ft. (no. 1067) ; southeast side, common at 450 ft. (no. 1066). NarsporoucH IsL.: south and east sides, Snodgrass and Heller; north side, common on lava beds near the shore (no. 1068). Srymour Ist., sourH: Snodgrass and Heller. ‘Tower Ist.: dried remains, the specific identity of which is somewhat in doubt, (no. 1069). WrENMaAN Ist.: common on tops of the cliffs. Possibly this is the species of which Mr. Heller noticed dried remains on the islet north of Wenman, mentioned by Robinson (1), 251. Endemic. C. polystachyus Roth. Desc. & Ic. 39, t. 11, f. 1 (1786). C. fugax Liebm. Mex. Halv. 8 (1850) ; Rob. (1), 127,—Cuatu- AM Ist.: Wreck Bay, Baur. Widely distributed. C. rotundus Hook. f. (3), 177; Rob. (1), 128.—ALBE- MARLE Ist.: Macrae. ‘The specific identity of the Macrae specimen is somewhat in doubt. Endemic? C. rubiginosus Hook. f. (3), 178; Rob. (1), 128.—CHarRLEs Ist.: Darwin. CuatHam Ist.: Wreck Bay, abundant in moist places, 450-700 ft., (no. 1070). Duncan Isi.: abund- ant among rocks at 1000 ft. (no. 1071). Endemic. Var. cornutus Rob. (1), 128. Mariscus cornutus Anderss. (1), 151: “Cy cormutus Anderss: \(2), 53, t. 13:4. 1-—Bar- RINGTON Is~.: common around dried pools (no. 1074). Cuar.es Isu.: Andersson. Duncan Ist.: Snodgrass and Heller. Srymour Ist., soutH: Snodgrass and Heller. En- demic. C. strigosus L. Sp. Pl. 47 (1753); Rob. (1), 128.— Cuar.es Ist.: Darwin. CHatHamM Isi.: Andersson, Fur- ther distr. U. S. C. suranimensis Rottb. Descr. Nov. Pl. 35, t. 6, f. 5 (1786) ; Rob. (1), 129—CuatuHam Ist.: Wreck Bay, in open grassy areas at 1750 ft. (no. 1075). James Ist.: Darwin. Further distr. SS. Us S:, Mex, Wind: S: Am: Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 43 C. tristachyus Boeck. Linnaea XXXV. 454 (1867-1868) ; Rob. (1), 129.—Cuaruam Ist.: Wreck Bay, Baur. Further distr. Mex., northern S. Am. C. sp.—BarRincton Isi.: dried remains of a Cyperus were found growing quite abundantly in coarse sandy soil in the vicinity of the shore. It seems to be different from any of the other species collected on the islands, (no. 1076). C. sp. Rob. (1), 129.—NarsoroucH Ist.: Snodgrass and Feller. C. sp. Rob. (1), 129-—Wenmawn Ist.: Snodgrass and Hel- ler. Probably C. Mutisii. Dichronema Michx. D. colorata (L.) Hitchk. Baham, in Mo. Bot. Gard. 1V. 141 (1893). Schoenus coloratus L. Sp. Pl. 41 (1753). D. leuco- cephala Michx. Fl. I. 37 (1803) ; Rob. (1), 129.—CuHatHam Ist.: Wreck Bay, covers the ground in dense mats in the open country around 650 ft. (no. 1077). Further distr. S. U. S., Mex., W. Ind., S. Am. Eleocharis R. Br. E. capitata (L.) R. Br. Prodr. 225 (1810). Scirpus capi- tatus L. Sp. Pl. 48 (1753).—ALBEMARLE Isx.: Villamil, occa- sional in rather loose dry soil on the south side of the rim of the crater at 3150 ft. (no. 1078). CHatTuHam Ist.: Wreck Bay, abundant in marshy ground around 1700 ft. (no. 1079). Fur- ther distr. tropical regions. E. fistulosa (Poir.) Schult. Mant. II. 89 (1824). Scirpus fistulosa Poir. Encycl. VI. 749 (1804). £. fistulosa Schult. 1. c.; Rob. (1), 129.—Cuatuam Ist.: Wreck Bay, common in pools of water at 1000-1750 ft. (no. 1080). Widely distrib- uted. E. mutata (L.) R. Br. Prodr. 224 (1810). Scirpus mutatus L. Syst. ed. 10, 867 (1759). E. mutata R. Br. 1. c.; Rob. (1), 129.— ALBEMARLE Ist.: Elizabeth Bay, Snodgrass and Hel- ler; Villamil, abundant. in brackish swamps and in other moist situations in the vicinity of the shore. The inhabitants of this 44 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. island use the dried stalks of this species for making pads for pack-saddles, and sleeping mats, (no. 1081). Further distr. U. S., Mex., W. Ind., S. Am. Fimbristylis Vahl F. capillaris (L.) A. Gray, Man. Bot. ed. 5, 567 (1869). Scirpus capillaris L. Sp. Pl. 49 (1753). F. capillaris A. Gray, l. c.; Rob. (1), 129.—AxBemarte Ist.: Tagus Cove, abund- ant in lava crevices, 500-4000 ft., (nos. 1083-1084) ; Villamil, occasional in lava crevices on the floor of the crater at 2750 ft. (no. 1082). Brnptok Ist.: Snodgrass and Heller. CHARLES IsL.: common among rocks near the shore (no. 1085). Nar- BOROUGH Ist.: Mangrove Point, Snodgrass and Heller. Widely distributed. F. diphylla (Rtz.) Vahl, Enum. II. 289 (1805). Scirpus diphyllus Rtz. Obs. V. 7 (1789). F. diphylla Vahl, 1. c.; Rob. (1), 129.—Hoop Ist.: Snodgrass and Heller. Widely dis- tributed in tropical regions. Hemicarpha Nees H. micrantha (Vahl) Britton, Bull. Torr. Club. XV. 104 (1888). Scirpus micranthus Vahl. Enum. II. 254 (1805). H. subsquarrosa Nees, in Mart. Fl. Bras. II. pt. 1, 61 (1824) ; Rob. (1), 130.—CuHatuHam Ist.: Wreck Bay, Baur. Further distr. U. S., Mex., W. Ind., S. Am. Kyllinga Rottb. K. pumila Michx. Fl. I. 28 (1803); Rob. (1), 130.— Cuar.es Ist.: in moist shady places around 1000 ft. (nos. 1086-1087). CHatTHAm Ist.: Wreck Bay, occasional in open country around 600 ft. (no. 1088). Further distr. Mex., W. Ind., S. Am. Scleria Berg. S. pterota Pres] in Oken, Isis, XXI. 269 (1828). S. praten- sis Lindl. ex Nees, Nov. Act. Nat. Cur. XIX. Suppl. I. 121 (1843) ; Rob. (1), 130.—CHatuHam Ist.: Wreck Bay, occa- sional patches at 650 ft. (no. 1089). Further distr. Mex., W. Ind., S. Am. Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 45 LEMNACEAE Lemna L. L. minor L. Sp. Pl. 970 (1753).—ALBEMar_e Ist.: Villa- mil, common in pools of slightly brackish water near sea level (no. 1100). CHatrHam Ist.: Wreck Bay, abundant on the surface of pools and streams, 1000-1800 ft., (no. 1101). Widely distributed. L. sp., Wolf (1), 284.—Cuartes Is. BROMELIACEAE Tillandsia L. T. insularis Mez in DC. Monog. IX. 756 (1896) ; Rob. (1), 130.—ALBEMARLE Ist.: Tagus Cove, rare above 2500 ft. on the west side of the mountain; on the southeast side it often covers the ground in great profusion over considerable areas; Villamil, common on the branches of trees and on the ground in vegetable mold, 350-1300 ft. CHarves IsL.: common on bushes, on small treés, among rocks in vegetable mold at 1400 ft. CHatHam Ist.: Wreck Bay, among rocks in vegetable mold, and covering the branches of Hippomane Mancinella trees around 700 ft. Duncan Isx.: on rocks and in vegetable mold, 1150-1250 ft. INDEFATIGABLE IsL.: Academy Bay, on the branches of trees and in vegetable mold, 350-550 ft., (no. 1119) ; northwest side, occasional at 550 ft., abundant in the region around 700 ft., where it often forms large patches on the ground in places where the vegetation is not too dense for its growth. James Ist.: James Bay, occasional on the sides of the bluffs, 1300-1500 ft. NaArsorouGH Isx.: south side, upper regions acc. to R. H. Beck. This is the only tank epiphyte found on the islands. Specimens often contain as much as a pint of water, from which they seem to obtain their entire sup- ply of moisture during dry weather. The root system is so poorly developed that a slight push will uproot a specimen when found growing on the ground. Endemic. COMMELINACEAE Commelina Plum. C. nudiflora L. Sp. Pl. 41 (1753); Rob. (1), 130.—Asinc- pon Ist.: common in the upper regions (no. 1122). ALBE- 46 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. MARLE Ist.: Iguana Cove, common on the sides of the cliffs above the cove (no. 1124); Tagus Cove, occasional on lava beds at 1400 ft. (no. 1123). CHartes Ist.: common at 1750 ft. (no. 1127). CHatruam Ist.: Wreck Bay, common at 350- 2100 ft. (nos. 1125-1126). Duncan Ist.: common on the sides of cliffs at 1250 ft. (no. 1128). INDEFATIGABLE IsL.: Academy Bay, abundant above 100 ft. (nos. 1130-1131); southeast side, occasional in open woods at 450 ft. (no. 1129). James Isu.: Darwin: Further distr. general in tropics. Commelinacea, Caruel (1), 621; Rob. (1), 131.—CHartes Ist.: Chierchia. Same as the preceding, acc. to Rob. 1. ¢. IRIDACEAE Iris 1. I. sp. ALBEMARLE IsL.: Villamil, a few sterile specimens of a small Jris were collected on the south rim of the crater at 3150 ft. (no. 1133). CANNACEAE Canna L. C. sp. ALBEMARLE Isi.: Villamil, occasional at 700 ft. In- DEFATIGABLE Ist.: Academy Bay, acc. to J. S. Hunter. JAMES Ist.: James Bay, occasional in woodland at 1500 ft. The specimens are all sterile. AMARYLLIDACEAE Furcraea Vent. F. cubensis Vent. in Bull. Soc. Philom. I. 66 (1793).— ALBEMARLE Ist.: Villamil, around habitations. CHARLES Ist.: around former habitations (no. 1134). CuHatuam IsL.: Wreck Bay, formerly used as hedges around the plantation. INDEFATIGABLE IsL.: northwest side, occurs first at 450 ft. and extends to above 1000 ft. This species was introduced on this island many years ago by the tortoise hunters who planted it at their plantation above 1000 ft. Capt. Thomas Levick, of Chatham Isl., told us that on one of his trips to Indefatigable, a few years ago, he took some of the seed with him and scat- tered it along the trail as he came down the side of the moun- Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 47 tain. From these seeds the plant has grown abundantly and now forms impenetrable thickets, many acres in extent, along the trail. The inhabitants of Chatham Isl. use the fiber of this plant for rope, of which it makes a very good quality. Widely distributed in tropical regions through cultivation. Probably introduced on the islands. Hypoxis L. H. decumbens L. Amoen. Acad. V. 396 (1759); Rob. (1), 131.—ALBEMARLE Ist.: Villamil, common in open woodland at 600 ft., rare at 1300 ft., (nos. 1135-1136). CHartes Ist. ; Darwin. CHATHAM Ist.: Wreck Bay, Baur. Further distr. Mex., W. Ind., S. Am. H. sp.—Asrncpon Ist.: an Hypo.vis, which is probably the last mentioned, occurs on this island. No specimens were taken. ORCHIDACEAE Epidendrum L. E. spicatum Hook. f. (3), 180; Rob. (1), 131—Asincpon Isu.: on the trunks and branches of trees around 1900 ft. (no. 1137). ALsBemarce Ist.: Villamil, abundant on the trunks and branches of trees, 1200-3150 ft., (no. 1138). CHARLES Ist.: Lee. James Ist.: James Bay, on trees above 2100 ft. (no. 1139). Endemic. Eulophia R. Br. E. sp.—INDEFATIGABLE Is~.: Academy Bay, a sterile specimen of an Orchid, with foliage similar to an Eulophia, was found growing in vegetable mold in densely shaded places at 600 ft. Hemsley, Gard. Chron. 177 (1900), refers to an Eulophia from the Galapagos Ids. It is possible that the speci- men Mr. Hemsley refers to and the one under consideration belong to the same species, (no. 1144). Ionopsis HBK. I. utricularioides (Sw.) Lindl. Coll. Bot. t. 39 A (1821- 1825). Epidendrum utricularioides Sw. Prodr. 122 (1788). —ALBEMARLE Ist.: Villamil, common on the trunks and 48 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. branches of trees, 300-700 ft., (no. 1141). Cuartes Ist.: fairly common on trees at 1300 ft. (no. 1142). Duncan Ist.: rare on bushes at 1250 ft. (no. 1144). INDEFATIGABLE IsL.: Academy Bay, on the branches of trees, 350-500 ft. ; northwest side, occasional at 800 ft.; southeast side, common on bushes and trees above 450 ft. (no. 1145). James Ist.: James Bay, occasional on the branches of trees at 1000 ft. It is not abund- ant and does not seem to extend above this elevation, (no. 1146). Further distr. W. Ind. Ponthieva R. Br. P. maculata Lindl. in Ann. & Mag. Nat. Hist. I. XV. 385 (1845 ).—James Ist.: James Bay, rare on the branches of trees in the upper moist regions. Specimen collected by Mr. R. H. Beck, identified by the late Mr. A. A. Eaton, (no. 1147). Further distr. Mex., northern S. Am. DICOTYLEDONEAE PIPERACEAE Peperomia R. & P. P. flagelliformis Hook. f. ex. Miq. in Hook. Lond. Jour. Bot. IV. 423 (1845), and (3), 181; Rob. (1), 131.—James Isz.: Darwin. Endemic. P. galapagensis Hook. f. ex. Miq. in Hook. Lond. Jour. Bot. IV. 426 (1845), and (3), 180; Rob. (1), 131.—Asincpon Ist.: occasional on trees at 1500 ft. (no. 1153). ALBEMARLE Ist.: Villamil, common on the branches of trees above 400 ft. (no. 1158). Duncan Isx.: on rocks and bushes at 1275 ft. (no. 1149). INDEFATIGABLE Ist.: Academy Bay, on rocks and trees above 350 ft. (no. 1152) ; southeast side, on trees and bushes at 625 ft. (no. 1151). James Ist.: James Bay, on the branches of trees above 1300 ft. (no. 1150). Specimens de- termined by Mr. Casimir de Candolle. Endemic. P. galioides HBK. Nov. Gen. & Sp. I. 71, t. 17 (1815) ; Rob. (1), 131.—Asrnepon IsL.: common in woodland at 1650 ft. (no. 1154). ALBEMARLE Ist.: Villamil, occasional in vegeta- ble mold among rocks, 1300-1500 ft., (no. 1156). CHaTHAM Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 49 Ist.: Wreck Bay, on rocks, 650-700 ft., (no. 1158). InpE- FATIGABLE Ist.: Academy Bay, occasional in vegetable mold among rocks in open woodland at 400 ft. (no. 1159) ; north- west side, common above 500 ft. (no. 1155). The specimens from Albemarle and Indefatigable Ids. were determined by Mr. Casimir de Candolle. Further distr. Mex., S. Am. P. obtusilimba C. DC. nov. sp. Foliis ternis-quaternis breviter petiolatis subovato-ellipticis basi et apice rotundatis utrinque glabris superne minutissime in margine ciliatis, 5-nerviis, nerviis tenuissimis; nervulo marginali obscuro ab apice fere usque ad medium decurrente; petiolo margine minute cilia- tis; spicis axillaribus terminalibusque, pedunculis minutissime puberulis petiolos superantibus; spicis ipsis limbos multo vel pluries superan- tibus filiformibus sublaxifloris, bractea orbiculari centro subsessili; ovario obovato emerso fere in apice stigmatifero, stigmate minuto glabro; bacca subovato-globosa glandulis asperulata. Caulis % mm. crassus minutissime puberulus. Limbi in sicco membranaceis rufes- centes epunctati, usque ad 10 mm. longi et 5 mm. lati. Petioli 14% mm. longi. Pedunculis usque ad 6 mm. longi. Spicae terminales 5 cm. axillares 2% cm. longae, % mm. crassae. Bractea diametro % mm. brevior. Bacca % mm. paululo longior. CuHaArLEs IsL.: common on rocks and low bushes at 1400 ft. (nos. 1160-1161). Endemic. P. petiolata Hook. f. (3), 181; Rob. (1), 131.—James Ist.: Darwin. Endemic. P. ramulosa Anderss. (1), 158, and (2), 57; Rob. (1), 131. —CHARLEs IsL.: common in decayed moss on the branches of trees at 1700 ft. (no. 1162). Endemic. P. Snodgrassii C. DC. in Rob. (1), 131—ALBeEMmarte Ist. : Iguana Cove, Snodgrass and Heller. Endemic. P. Stewartii C. DC. nov. sp. Foliis sat longe petiolatis oblongo-ovatis basi acutis apice subacutis, utrinque glabris superne margine ciliolatis, 6-nerviis, nerviis tenuis- simis; petiolo margine crispulo-hirtello; pedunculis terminalibus gla- bris petiolos fere aequantibus; spicis folia pluries superantibus sub densifloris; bractea obovata supra centrum longiuscule pedicellata; antheris rotundatis quam filamenta multo brevioribus; ovario emerso ovato apice obtuso, stigmate puberulo; bacca ovata glandulis globosis asperata. Caulis crispulo-hirtellus filiformis fere 34 mm. crassus. Folia alterna, internodia 6-7 mm. longa. Limbi in sicco membranaceli, superi 15-18 mm. longi et 8-9 mm. lati, inferi magis ovati. Petioli circiter 6 mm. longi. Spicae circiter 5 cm. longae et 1 mm. crassae. Bracteae infimae orbiculares, aliae ut in diagnosi. Ovarium paullo sub apice stigmatiferum. Bacca 1 mm. longa, sessilis. 50 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. ABINGDON IsL.: common on rocks at 1050 ft. (no. 1163). ALBEMARLE Ist.: Villamil, common among rocks in woodland, 350-1500 ft., (nos. 1164-1165). CHartes IsL.: in moist shady places at 1000 ft. (no. 1166). INDEFATIGABLE IsL.: Academy Bay, in shady places, 50-400 ft., (no. 1168, Type) ; northwest side, common in shade at 950 ft. (no. 1170). James Ist.: James Bay, in woodland at 850 ft., not common, (no. 1171). This is one of the most common species of Peperomia found on the islands and is usually the first species to be seen in ascending the sides of the mountains. Endemic. P. n. sp. Rob. (1), 132.—AtLsBemarte Ist.: Tagus Cove, Snodgrass and Heller. Endemic. P. sp. Rob. & Greenm. (1), 148.—Cuatuam Ist.: Baur. URTICACEAE Fleurya Gaud. F. aestuans Gaud. in Freyc. Voy. Bot. 497 (1826); Rob. (1), 132.—Asrnepon Ist. : common, 800-1100 ft., (no. 1172). ALBEMARLE IsL.: Iguana Cove, abundant from the beach to 600 ft. (no. 1325); Tagus Cove, common in lava crevices in shade, 300-2900 ft., (no. 1177); Villamil, abundant among rocks, 300-1300 ft., (nos. 1174-1175). CHarves IsL.: occa- sional among rocks at 1550 ft. (no. 1326). CHaTHAmIsL.: Basso Point, occasional in shady places at 900 ft. (no. 1327). Duncan Ist.: common on the sides of steep lava cliffs at 1000 ft., also common around 1250 ft., (no. 1328). Hoop Ist.: occasional in lava crevices, 400-600 ft., (no. 1329). InpDEFAr- IGABLE Ist.: Academy Bay, occasional among rocks at 50 ft. At this elevation the specimens are low and with many stinging hairs on the stem. This same species also grows very abund- antly around 600 ft., where it attains a height of 3-4 ft. and has fewer stinging hairs on both the stems and leaves than do the specimens taken from the lower elevations, (nos. 1130- 1331). James Ist.: James Bay, Snodgrass and Heller. Nar- BOROUGH IsL.: south side, Snodgrass and Heller. This species shows much variation both in size and in the arming of the stem and leaves, but the differences are not sufficient to be of formal value. Further distr. Mex., W. Ind., S. Am. Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 51 Parietaria L. P. debilis G. Forst. Fl. Ins. Aust. Prodr. 73 (1786) ; Rob. (1), 132.—AxBemarte IsL.: Iguana Cove, in protected places on the sides of the cliffs above the cove (no. 1334); Tagus Cove, occasional at 1000 ft., common in lava crevices at 2850 ft., (no. 1333) ; Villamil, common among rocks at 550 ft. (no. 1332). CHarzes IsL.: in shady places among rocks at 1550 ft. (no. 1335). James Ist.: Darwin. Widely distributed in tropical regions. Pilea Lindl. P. Baurii Rob. (1), 133——Apinepon Ist.: common in moist shady places around 1650 ft. (no. 1336). CHARLES Ist.: Baur. CHATHAM Ist.: Wreck Bay, common in open country around 1000 ft. (no. 1338). James Ist.: James Bay, occasional around 2000 ft. (no. 1339). There is much varia- tion in the specimens found growing in sun and in shade, those growing in the shade having a green stem, thinner leaves, and a much less branched inflorescence. Endemic. P. microphylla (L.) Liebm. in Vidensk. Selk. Skr. ser. 5, IT. 296 (1851). Parietaria microphylla L. Sp. Pl. ed. 2, 1492 (1763). Pilea muscosa Lindl. Coll. Bot. t. 4 (1821); Rob. (1), 133.—James Ist.: Darwin. Further distr. Mex., W. Ind., S. Am. P. peploides (Gaud.) Hook & Arn. Bot. Beech. 96 (1832). Dubreulia peploides Gaud. in Freye. Voy. Bot. 495 (1826). P. peploides Hook & Arn. 1. c.; Rob. (1), 133.—ALBEMARLE Ist.: Tagus Cove, in lava crevices at 2850 ft. (no. 1340). CHARLES IsL.: common on moist rocks at 1000 ft., occasional at 1550 ft., (nos. 1339, 1341). CHatHAm Ist.: Wreck Bay, rare in shady places at 700 ft. (no. 1342). James Ist.: Dar- win. Further distr. Pacific Ids., Asia. Urera Gaud. U. alceaefolia (Poir.) Gaud. in Freyec. Voy. Bot. 497 (1826). Urtica alceaefolia Poir. Suppl. 227 (1816).—ALBE- MARLE Isi.: Villamil, common bushes, 650-1500 ft. The leaves of many of the specimens are variegated, (no. 1343). INDEFATIGABLE Is~.: Academy Bay, common bushes above 52 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. 500 ft. They increase in size and abundance above this eleva- tion, according to Mr. F. X. Williams, of the Academy’s expe- dition, (no. 1344). James Ist.: James Bay, occasional bushes in open woodland around 2000 ft. (no. 1345). Further distr. Mex., S. Am. LORANTHACEAE Phoradendron Nutt. P. florianum (Anderss.) Rob. (1), 133. Viscum florianum Anderss. (1), 219, (2), 92.—Cwartes Isu.: Andersson. En- demic. P. galapageium (Hook. f.) Rob. (1), 133. Viscum galapa- geium Hook. f. (3), 216—CuatHam Ist.: Darwin; Anders- son. Endemic. P. Henslovii (Hook. f.) Rob. (1), 133. Viscum Henslovit Hook. f. (3), 216—Apsrtncpon IsL.: common on trees and bushes, 450-1000 ft., (no. 1102). ALBEMARLE IsL.: Cape Rose, on trees near the-shore (no. 1103); Cowley Bay, on trees and bushes above 400 ft. (no. 1106) ; Iguana Cove, com- mon on trees above 300 ft. (no. 1104); Tagus Cove, on trees and bushes, 400-4000 ft. ; Villamil, common on bushes near the shore. It also occurs throughout the wooded regions to 1500 ft. and is present on small trees and bushes on the rim of the crater at 3150 ft., as well as on trees of Zanthoxylum Fagara on the floor of the same at 2750 ft., (no. 1105). CHaRLEs IsL.: common on bushes of Lipochaeta laricifolia, 600-1000 ft., also common on trees of Zanthoxylum Fagara and Scalesia pedunculata at 1100 ft., (no. 1108). CHatHam Ist.: Basso Point, occasional on trees at 900 ft. (no. 1107) ; Sappho Cove, on trees and bushes near the shore; Wreck Bay, Baur. Dun- cAN IsL.: occasional on bushes at 1250 ft. (no. 1109). INpDE- FATIGABLE Ist.: Academy Bay, abundant on trees and bushes near the shore. It increases in size and abundance with the elevation above sea level, (no. 1102) ; southeast side, common on trees and bushes above 400 ft.; northwest side, abundant above 700 ft. James Ist.: James Bay, abundant on bushes to 2500 ft.; northeast side, on trees and bushes above 100 ft. Jervis Ist.: occasional on bushes above 700 ft. This species Vor. I) STEWART—BOTANY OF THE GALAPAGOS ISLANDS 53 varies greatly in size at different elevations. Specimens from the moist region are usually much larger than those found in the dry and transition regions. Endemic. P. uncinatum Rob. (1), 134—Narporoucu Ist.: Snod- grass and Heller. Endemic. POLYGONACEAE Polygonum L. P. acre HBK. Nov. Gen. & Sp. IT. 179 (1817). —CuatnHam Ist.: Wreck Bay, common in pools of water at 1000 ft. (no. 1121). Further distr. U. S., Mex., W. Ind., S. Am. P. acuminatum HBK. Nov. Gen. & Sp. IT. 178 (1817) ; Rob. (1), 134—Garapacos Ips.: according to Griesb. Fl. W. Ind. 161. It is probable that the next species has been mistaken for this one, as the two resemble each other rather closely. Further distr. Mex., W. Ind., S. Am. P. galapagense Caruel (1), 624; Rob. (1), 134.—Atpr- MARLE Isi.: Villamil, occasional above 2500 ft. CuHatHam Ist.: Wreck Bay, common in large bunches 2-4 ft. high in the open grassy country above 1700 ft. (no. 1120). Endemic. CHENOPODIACEAE Atriplex L. A. sp. Rob. (1), 134.—InpEraticaBLe Ist.: north side, low shrubs on sand beaches (no. 1346). Srymour IsL., NorTH: Snodgrass and Heller. All of the specimens are sterile and in- determinate as to species. A. sp. Rob. (1), 134—WeEnman Ist.: Snodgrass and Heller. Salicornia L. S. sp. (?).—JaAmes Is.: northeast side, a plant resembling a Salicornia in habit and inflorescence was seen growing on the shores of salt lagoons. No specimens were secured. 54 : CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. AMARANTACEAE Alternanthera Forsk. A. radicata Hook. f. (4), 261, 262; Rob. (1), 134.— CuatHam Isi.: Darwin. CHARLES IsL.: abundant in barren places among lava boulders near the shore (no. 1347). Hoop Ist.: Snodgrass and Heller. Endemic. A. rigida Rob. & Greenm. (1), 143, 148; Rob. (1), 135.— James Ist.: northeast side, occasional bushes 6-10 inches high on lava beds near the shore, and to some extent at 700 ft., (no. 1348). Endemic. A. subscaposa Hook. f. (3), 189; Rob. (1), 135.—CHarLEs Ist.: Darwin. DuNcAN IsL.: rare in moist protected places around 1250 ft. (no. 1349). Endemic. Amaranthus L. A. caracasanus HBK. Noy. Gen. & Sp. IT. 195 (1817) ; Rob. (1), 135.—ALBEMARLE IsL.: Iguana Cove, rare on the sides of the cliffs above the cove (no. 1351); Tagus Cove, Snod- grass and Heller; Villamil, fairly common in open places on the lower parts (no. 1350). CHarves Ist.: abundant from the beach to 1000 ft. during the rainy season (no. 1354). CuatHam Ist.: Wreck Bay, abundant near the shore (no. 1356). INDEFATIGABLE IsL.: northwest side, Andersson. One of the common spring weeds of the islands where it occurs. Further distr. northern S. Am. A. celosioides HBK. Nov. Gen. & Sp. IL. 194 (1817) ; Rob. (1), 135.—Cuartes Ist.: Darwin; Andersson. CHATHAM Ist.: Andersson. Further distr. northern S. Am. A. sclerantoides Anderss. (2), 59, t. 2, f. 1; Rob. (1), 135. —BarrincTon Ist.: Snodgrass and Heller. CHarwes Ist.: common in open sunny places at 450 ft. (no. 1357). CHATHAM Ist.: Wreck Bay, common near the shore (no. 1358). Nar- BOROUGH IsL.: east side, Snodgrass and Heller. Endemic. Forma abingdonensis nov. forma. Ramulis diffusis; foliis linearibus late patentibus circa 2.5 cm. longis, ad apicem 1 mm. latis. ABINGDON IsL.: occasional among rocks at 700 ft. (no. 1359). Plate II, fig. 1. Endemic. Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 55 Forma albemarlensis nov. forma. Foliis subappressis ad apices dilatis 1.9 cm. longis, 5 mm. latis. ALBEMARLE Ist.: Turtle Cove, common on sand beaches, (no. 1360). Plate II, fig. 2. Endemic. Forma chathamensis Rob. & Greenm. (1), 140; Rob. (1), 135.—CuatHam Ist.: Wreck Bay, Baur. Endemic. Forma hoodensis Rob. & Greenm. (1), 140; Rob. (1), 135. —GARDNER Ist. (near Hoop Ist.) : common on sand beaches (no. 1361). Hoop Ist.: Baur; Snodgrass and Heller. En- demic. A. spinosus L. Sp. Pl. 991 (1753); Rob. (1), 135.— Cuartes Ist.: Andersson. Of wide distribution. A. squamulatus (Anderss.) Rob. Proc. Am. Acad. XLIII. 22 (1907). Scleropus squamulatus Anderss. (1), 162, (2), 60. A. squarrulosus Uline & Bray, Bot. Gaz. XIX. 270 (1894); Rob. (1), 135.—ALBemar_eE IsLt.: Cowley Bay, occasional at 2000 ft. (no. 1363) ; Tagus Cove, occasional in tufaceous soil at 100 ft. (no. 1362). Cartes Ist.: Snod- grass and Heller. CuatHam Ist.: Andersson. DUNCAN Ist.: Snodgrass and Heller. INDEFATIGABLE IsL.: Academy Bay, (no. 1364) ; north side, Snodgrass and Heller; northeast side, fairly abundant in loose ashy soil near the shore (no. 1365). James Ist.: James Bay, fairly common in rocky soil on the lower parts (no. 1356). Jervis Ist.: Bawr. SEYMOUR IsL., NortTH: Snodgrass and Heller. Endemic. A. urceolatus Benth. Bot. Sulph. 158 (1844); Rob. (1), 136.—INDEFATIGABLE IsL.: Andersson. Further distr. adja- cent S. Am. from Peru northward. Lower California acc. to Roba sc A. viridis Ie Sp. Pl. ed. 2, 1405 (1763); Rob. (1), 136.— ALBEMARLE Ist.: Villamil, occasional to 650 ft., abundant at 1300 ft., (mos. 1357-1359). Barrineton Ist.: Snodgrass and Heller. CHATHAM IsL.: Wreck Bay, abundant in sandy soil near the shore and in clay soil in open sunny places at 50 ft. (nos. 1360-1361). Further distr. general in warm coun- tries. January 11, 1911 56 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. Froelichia Moench. F. juncea Rob. & Greenm. (1), 143, 148; Rob. (1), 136.— ALBEMARLE Ist.: Elizabeth Bay, Snodgrass and Heller; Tagus Cove, Snodgrass and Heller; Villamil, in dense some- what prostrate clumps on lava beds near sea level (no. 1362). INDEFATIGABLE IsL.: southeast side, low bushes at 450 ft. (no. 1363). Endemic. F. lanigera Anderss. (2), 63; Rob. (1), 136. F. lanata Anderss. (2), t. 3, f. 1 ALBEMARLE IsL.: Cowley Bay, pros- trate bushes in pumice soil near the shore (no. 1364); Tagus Cove, abundant on the sides and top of the mountain at 4000 ft. (no. 1365). Duncan Ist.: low bushes on the sides and top of the island (no. 1366). NaArsoroucu Ist.: north side, low bushes on recent lava. Endemic. F. nudicaulis Hook. f. (3), 192; Rob. (1), 136.—CHARLEs Ist.: Darwin; Andersson. CHATHAM IsL.: Andersson. En- demic. F. scoparia Rob. (1), 136.—JAmes Ist.: James Bay, com- mon bushes on the lower parts (no. 1367). NARBorouUGH Ist.: south side, Snodgrass and Heller. Endemic. Tresine L. I. Edmonstonei Hook. f. (3), 190; Rob. (1), 137.— Cuartes Isu.: Darwin. Endemic. Pleuropetalum Hook. f. P. Darwinii Hook. f. (1), t. 2, (3), 221; Rob. (1), 137.— ALBEMARLE Ist.: Iguana Cove, Snodgrass and Heller; Villa- mil, occasional bushes 2-3 ft. high in woodland at 400-700 ft. (no. 1358). James Ist.: James Bay, common bushes in woodland above 1500 ft. (nos. 1369-1370). Endemic. Telanthera R. Br. T. echinocephala (Hook. f.) Mogq.-Tand in DC. Prodr. XIII. pt. 2, 373 (1849). Brandesia echinocephala Hook. f. (3), 189. T. echinocephala Mogq.-Tand. 1. c.; Rob. (1), 137.— ABINGDON IsL.: common in thickets of Laguncularia race- Vor.1] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 57 mosa near the shore, common bushes at 1000 ft. The speci- mens taken from the vicinity of the shore have much smaller leaves than do those taken at 1000 ft., (nos. 1371-1372). ALBEMARLE IsL.: Cowley Bay, common bushes at 2100 ft. (no. 1376); Iguana Cove, bushes 3-4 ft. high, all over the lower parts, (no. 1374); Villamil, common bushes to 600 ft. (no. 1375). Barrincton Ist.: Snodgrass and Heller. Cuar.es Ist.: Darwin; A. Agassiz; Andersson; Snodgrass and Heller. CHATHAM Ist,: Wreck Bay, occasional bushes 4-5 ft. high to 250 ft. (no. 1377). Duncan Ist.: A. Agassiz; Baur; Snodgrass and Heller. GarpNER Ist. (near Hoop Ist.): Snodgrass and Heller. Hoop Ist.: common bushes (nos. 1378-1380). INDEFATIGABLE IsL.: Academy Bay, com- mon bushes 5-6 ft. high in the vicinity of the shore, occasional to 300 ft., (no. 1381); north side, Snodgrass and Heller; southeast side, common bushes on the lower parts (no. 1382). James Ist.: James Bay, common bushes to above 1000 ft. (nos. 1383-1384). Endemic. T. filifolia (Hook. f.) Moq.-Tand. in DC. Prodr. XIII. pt. 2, 368 (1849). Bucholtzia filifolia Hook. f. (3), 192. T. filifolia Mogq.-Tand. 1. c.; Rob. (1), 138.—James Isu.: Scouler. En- demic. T. flavicoma Anderss. (1), 166, (2), 61, t. 5, £.2; Rob. (1), 138.—ApincGpon IsL.: prostrate bushes, common, 900-1400 ft., (nos. 1386-1387). ALBEMARLE IsL.: Cowley Bay, occa- sional bushes at 2000 ft. (no. 1393) ; Villamil, species in doubt (no. 1392). Cuarues Ist.: Andersson. CHATHAM ISL.: Basso Point, low bushes in open places at 875 ft. (no. 1388) ; Sappho Cove, occasional on lava flows and in the vicinity of the coast (no. 1389) ; Wreck Bay, low bushes near the beach (no. 1390). GarpNER Is~. (near Hoop Ist.): Snodgrass and Heller. Woop Ist.: abundant in crevices of the lava (no. 1391). INDEFATIGABLE Ist.: Academy Bay, occasional low bushes near the beach and in the interior to 100 ft. (nos. 1394- 1395). James Ist.: James Bay, occasional bushes 12-18 inches high to 2150 ft. (nos. 1396-1397). Endemic. T. galapagensis nov. sp. Suffrutescens circa 2.5 dm. alta; ramis oppositis vel alternis tere- tibus striatis glaucescentibus ad nodos lanuginoso-ciliatis; foliis oppo- sitis 1.2-2.8 cm. longis, 0.6-1.3 cm. latis, oblanceolatis apice obtusis basi 58 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. cuneatis sessilibus vel brevi-petiolatis integerrimis glaberrimis glau- cescentibus; spicis densifloris 3-9 mm. longis terminalibus et axillaribus sessilibus; bracteis ovatis carinatis acutis hispidis; sepalis exterioribus lanceolatis plerumque 3-costatis bruneis hispidis apice subflavis; sepalis interioribus lineari-lanceolatis acutis carinatis margine hyalinis; stam- inodiis ad apices laciniatis elongatis. This species is closely related to T. Snodgrassti Rob. but dif- fers in the glabrous glaucous character of the leaves and in the smaller size of the spikes. GarpNeER IsL. (near CHARLES Isu.): (no. 1403). J. R. Slevin, collector. Plate II, figs. 3-4. Endemic. T. glaucescens (Hook. f.) Mogq.-Tand. in DC. Prodr. XIII. pt. 2, 369 (1849). Bucholtzia glaucescens Hook. f. (3), 191. T. glaucescens Mogq.-Tand. 1. c.; Rob. (1), 138.—CHARLES Ist.: Andersson. CHATHAM IsL.: Darwin; Andersson. En- demic. T. halimifolia (Lam.) n. comb. Achyranthes halimifolia Lam. Dict. I. 547 (1783). T. frutescens Mog.-Tand. in DC. Prodr. XIII. pt. 2, 365 (1849) ; Rob. (1), 138—ALBEMARLE Ist.: Villamil, common in moist places in the upper regions, especially in open woodland around 1300 ft. (no. 1398). Cuatuam Ist.: Wreck Bay, common in woodland above 200 ft. (no. 1399). INDEFATIGABLE IsL.: Academy Bay, small specimens of this species occur at 300 ft., but at 600 ft. the specimens are larger and more abundant, (no. 1401); north- west side, common at 400 ft. (no. 1400); southeast side, abundant in shady places at 625 ft. (no. 1402). Further distr. S. Am. T. Helleri Rob. (1), 138—CuLpeppeEr Ist.: low bushes among rocks near the shore (no. 1404). F. X. Williams, col- lector. Endemic. Var. obtusior Rob. (1), 139—WeENMAN IsL.: common bushes, 2-3 ft. high, (no. 1424). oe nudicaulis (Hook. f.) Mogq.-Tand. in DC. Prodr. XIII. pt. 2, 369 (1849). Bucholtzia nudicaulis Hook. f. (3), 191. ile pudicelic Mog.-Tand. 1. c.; Rob. (1), 139. Ist.: common bushes on the lower parts (no. 1405). ALBE- MARLE Ist.: Cowley Bay, Baur; Tagus Cove, common bushes at 4000 ft. (no. 1407) ; Villamil, low and somewhat prostrate bushes (no. 1406). Brattre Ist.: (no. 1408). CHARLES Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 59 IsL.: common bushes (no. 1409). CHatHam Ist.: north side, Bawr. Duncan Ist.: Snodgrass and Heller. Hoon IsL.: species in doubt (no. 1411). InpeEFaticaBLeE Ist. : northeast side, occasional low shrubs (no. 1410). Jervis Is. : occasional low bushes at 1050 ft. (no. 1412). James Ist: James Bay, fairly abundant to 1200 ft. (no. 1413); Orchilla Bay, Baur. Further distr. S. Chili. T. rugulosa Rob. (1), 139—CuHatuHam Ist.: Wreck Bay, occasional low trees 10-12 ft. high around 1800 ft. (no. 1414). Endemic. T. Snodgrassii Rob. (1), 140.—ALsBemarte Isi.: Villamil, low bushes, fairly common at 550 ft., (no. 1415). James Isx.: James Bay, (no. 1416). Srymour IsL., NortTH: Snodgrass and Heller. Endemic. T. strictiuscula Anderss. (1), 166; Rob. (1), 140.—ALBE- MARLE IsL.: Iguana Cove, common bushes near the shore (no. 1418) ; Villamil, low bushes at 550 ft. (no. 1417). Caries Ist.: bushes 2-3 ft. high among rocks at 1400 ft. (no. 1420). CHATHAM Ist.: Wreck Bay, common bushes 3-6 ft. high at 500 ft. (no. 1419). INDEFATIGABLE IsL.: southeast side, low bushes at 600 ft. (no. 1421). Narsoroucu Ist.: south side, Snodgrass and Heller. Endemic. T. vestita Anderss. (1), 169, (2), 63, t. 4, f. 1; Rob. (1), 140.—INDEFATIGABLE Ist.: Academy Bay, common bushes on the lower parts (no. 1423) ; north side, common in lava crev- ices (no. 1422). Endemic. BATIDACEAE Batis L. B. maritima L. Syst Nat. ed. 10, 1376 (1760); Rob. (1), 141.—Cuar_es Ist.: common on sand beaches (no. 1425). CuatHam IsL.: Sappho Cove, common near the shore (no. 1426). INDEFATIGABLE IsL.: north side, common on sand beaches (no. 1427); southeast side, common on sand beaches (no. 1428). James Ist.: James Bay, common around salt lagoons and around the borders of a crater lake south of the bay (no. 1430). Widely distributed on tropical shores. 60 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. BASELLACEAE Boussingaultia HBK. B. baselloides HBK. Nov. Gen. & Sp. VII. 196, t. 645 (1825); Rob. (1), 141.—Cuartes Ist.: Darwin. DuNcAN Is_.: trailing vines covering rocks at 1150 ft. (no. 1431). Further distr. Mex., W. Ind., S. Am. PHY TOLACCACEA Phytolacca L. P. octandra L. Sp. Pl. ed. 2, 631 (1762). P. decandra Hook. f. (3), 193, not L.; Anderss. (1), 227, (2), 97; Rob. (1), 141.—ALBEMARLE Ist.: Villamil, prostrate bushes, com- mon around 3150 ft., (no. 1433). James Ist.: James Bay, common bushes 4-5 ft. high above 2150 ft. (no. 1432). Rob- inson, |. c., suggested that this species might possibly be P. octandra. The specimens secured confirm this suggestion. Further distr. Mex., W. Ind., S. Am. Rivina Plum. R. humilis L. Sp. Pl. 121 (1753).—Atpemarte Ist.: Villa- mil, low bushes in dense woodland at 500 ft. (no. 1434). INDEFATIGABLE IsL.: Academy Bay, rare near sea level, com- mon in dense woodland at 300-450 ft., (nos. 1435-1436). James Ist.: James Bay, fairly abundant in woodland around 2100 ft. (no. 1437). Further distr. S. U. S., Mex., W. Ind., S. Am: NYCTAGINACEAE Boerhaavia L. B. erecta L. Sp. Pl. 3 (1753); Rob. (1), 141.—ALBEMARLE Ist.: Macrae. CuHartes Is~.: common to 600 ft. during the spring months; during the autumn it was found occasionally among rocks at 1450 ft., (no. 1438). CHaTHAM Ist.: Andersson. INDEFATIGABLE IsL.: Andersson. Further distr. S. U.S. Mex., W. Ind.,S: Am- B. paniculata Rich. Act. Soc. Nat. Hist. Par. I. 105 (1792) ; Rob. (1), 141.—ALBEMARLE Ist.: Tagus Cove, common in Vor.1]- STEWART—BOTANY OF THE GALAPAGOS ISLANDS 61 open sunny places in tufaceous soil on the lower parts (no. 1440). James Ist.: Darwin. NarsorouGH Ist.: north side, occasional in lava crevices (no. 1441). Further distr. S. U. S., Mex., S. Am. B. scandens L. Sp. Pl. 3 (1753); Rob. (1), 141.—ALBE- MARLE Ist.: Iguana Cove, abundant near the shore (no. 1443) ; Villamil, abundant in open places on the lower parts of the island (no. 1442). CHAR Es IsL.: common in open grassy places around 1000 ft. (no. 1444). CuatHam Isi.: Anders- son; Snodgrass and Heller. DUNCAN Isu.: occasional among bushes at 1150 ft. (no. 1466). INDEFATIGABLE IsL.: north- west side, common to 800 ft., very abundant in woodland around 650 ft., where it often forms the principal undergrowth, (no. 1445); southeast side, fairly common at 600 ft. (no. 1446). James Ist,: James Bay, common in open woods at 850 ft. (no. 1447). Further distr. S. U. S., Mex., W. Ind., Sem! B. viscosa Lag. & Rod. Anal. Cienc. Nat. TV. 256 (1801) ; Rob. (1), 142.—Apsrnepon Ist.: common on lava beds near the shore (no. 1448). ALBEMARLE Ist.: Tagus Cove, com- mon in open sunny places in tufaceous soil to 1000 ft. (no. 1449) ; Villamil, abundant in light ashy soil and on lava beds on the lower parts (no. 1450). Brarttie Ist.: (no. 1455). CHARLES IsL.: common in tufaceous soil to 650 ft. (no. 1451). CuHaTHAM IsL.: Basso Point, common on sand beaches (no. 1454). GarpNner Ist. (near Hoop Ist.): Snodgrass and Heller. Hoop Ist.: common on hillsides at 250 ft. (no. 1456). INDEFATIGABLE IsL.: north side, Snodgrass and Heller; south- east side, common in tufaceous soil at 600 ft. (no. 1458). James Ist.: James Bay, Snodgrass and Heller. A species which is rather characteristic of open sunny places in the dry region. Further distr. S. W. U. S., Mex., W. Ind., S. Am. Cryptocarpus HBK. C. pyriformis HBK. Nov. Gen. & Sp. I. 188, t. 124 (1817) ; Rob. (1), 142.—Asincpon Ist.: forming low thickets on sand beaches (no. 1458). ALBEMARLE IsL.: Elizabeth Bay, Snodgrass and Heller; \guana Cove, in dense thickets near the shore; Tagus Cove, Snodgrass and Heller; Turtle Cove, cover- 62 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. ing large areas with a dense growth of low bushes in the vicinity of the shore; Villamil, common on sand beaches and to some extent in the interior around brackish water pools (no. 1459). Barrincton Ist.: forming low thickets on sand beaches (no. 1462). BinpLor Ist.: Snodgrass and Heller. Occasional thickets of this species were noticed at various places along the north shore. CHARLES IsL.: bushes 3-6 ft. high, forming tangled thickets on sand beaches, (no. 1460). CHATHAM Ist.: Basso Point, on sand beaches and on lava flows in the interior; Sappho Cove, bushes on the beach and in the interior (no. 1461) ; Wreck Bay, fairly common near the shore. Duncan Ist.: bushes near the shore. GARDNER IsL. (near Hoop Ist.): low bushes on the beach. Hoop Ist.: very abundant in dense low thickets on sand beaches, and to some extent in the interior at 600 ft., (no. 1463). InpErat- IGABLE Ist.: Academy Bay, common on the beach and occa- sional at various places in the lower dry region; north side, low bushes on the beach; southeast side, common in thickets among rocks and in sand (no. 1465). It was also noticed in various other places on the shore, while the “Academy” was sailing around the island. JAmes Ist.: James Bay, common bushes on the beach and around the shores of salt water lagoons (no. 1456). Jervis Ist.: low bushes on the beach. NARBOROUGH IsL.: east side, Snodgrass and Heller. StyMour IsL., SOUTH: occasional on the beach and in thickets of Discaria pauciflora and Maytenus obovata bushes. Further distr. Ecuador, Bo- livia. Mirabilis L. M. Jalapa L. Sp. Pl. 177 (1753).—ALBeMar_LE Ist.: Villa- mil, in gardens, and undoubtedly introduced, (no. 1459). Pisonia L. P. floribunda Hook. f. (3), 193; Rob. (1), 143.—Absinc- pon Ist.: common trees, 450-1650 ft., (no. 1460). ALBE- MARLE Ist.: Cowley Bay, common trees above 1300 ft.; Iguana Cove, Snodgrass and Heller; Tagus Cove, forest trees above 1500 ft.; Villamil, large trees, 100-900 ft. CHARLES Isy.: trees 10-30 ft. high, occasional around 1000 ft., (no. 1461). Duncan Ist.: low trees and bushes around 1150 ft. Vor. I) STEWART—BOTANY OF THE GALAPAGOS ISLANDS 63 (no. 1463). INDEFATIGABLE IsL.: Academy Bay, small trees near the shore, one of the common forest trees above 350 ft., (no. 1463); north side, trees above 1000 ft.; northwest side, small trees at 150 ft., forming large forest trees at 500-800 ft. James Ist.: James Bay, common trees, 450-1700 ft., (nos. 1464-1465) ; northeast side, trees above 600 ft. This species forms one of the most common forest trees in the transition and moist regions on the islands where it occurs. Hooker, op. c. 194, describes it as an almost leafless tree, but we found it to be usually covered with a dense growth of leaves. As a rule the trunk is short and the branches are large and broadly spread- ing. Owing to the rough nature of the bark it is usually cov- ered with epiphytes when it occurs in the moist regions. En- demic. AIZOACEAE Mollugo L. M. flavescens Anderss. (1), 226, (2), 96, t. 15, f. 2; Rob. (1), 143—ALBEMarRLE Ist.: Darwin; Macrae; Baur. Cuarces Ist.: Snodgrass and Heller. CHatHam Ist.: Sappho Cove, in lava crevices at 800 ft. (no. 1466); Wreck Bay, Baur. INDEFATIGABLE IsL.: north side, Snodgrass and Heller. Endemic. ‘ Var. floriana Rob. (1), 143——Cuar es Ist.: Cormorant Bay, abundant in coarse gravelly soil near the shore (no. 1467). Endemic. M. gracillima Anderss. (1), 226, (2), 96; Rob. (1), 143.— ALBEMARLE Ist.: Iguana Cove, occasional in ashy soil on sides of the cliff above the cove (no. 1468); Tagus Cove, abundant in tufaceous soil (no. 1470); Villamil, common in open places at 550 ft. (no. 1469). Brattre Ist.: (no. 1471). Cuar es Is_.: common in open grassy areas at 600 ft. The specimens taken here are very small, but they seem to possess the characters of this species, (no. 1472). CHATHAM Ist.: Basso Point, occasional in lava crevices (no. 1473). Duncan IsL.: common in dry places near the shore (no. 1474). JAmeEs Iszt.: Orchilla Bay, Baur. Narporoucu Isi.: north side, common on lava beds (no. 1476). WerNnMAN Ist.: (no. 1477). Endemic. 64 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. M. Snodgrassii Rob. (1), 144—ALBEMaRLE IsL.: Cowley Bay, bushes 1 ft. and more in height, rare in pumice soil, (no. 1478); Elizabeth Bay, Snodgrass and Heller; Tagus Cove, Snodgrass and Heller. NarsorouGH Ist.: Mangrove Point, Snodgrass and Heller. Endemic. Sesuvium L. S. Edmonstonei Hook. f. (3), 221; Rob. (1), 144.—Bar- RINGTON IsL.: common covering rocks along the shore (no. 1479). Brartre Ist.: (no. 1480). CHartves Ist.: common on sand beaches, forming bright red patches when seen from a distance, (no. 1481). CuLPEpPER IsLt.: common on the sides of cliffs. Duncan IsLt.: common among rocks along the shore and in occasional patches up to 250 ft. (no. 1482). GarDNER Ist. (near Hoop Ist.): Snodgrass and Heller. Hoop Ist.: very abundant on the tops of the cliffs at the east end of the island (no. 1483). The stems and leaves of this plant are usually bright red when it grows in open sunny places, but are green with but a small amount of the red color when it grows in the shade. Endemic. S. Portulacastrum L. Syst. Nat. ed. 10, 1058 (1760) ; Rob. (1), 144.—Absincpon IsL.: occasional on sand beaches. At- BEMARLE Ist.: Elizabeth Bay, Snodgrass and Heller; Turtle Cove, abundant on sand beaches (no. 1484). BArrincTon Ist.: on sand beaches. CHARLES Is~.: forms thick mats on sand beaches. It also occurs around the shores of salt lagoons where the water is saturated with salt. In such situations the leaves are somewhat reduced in size, (no. 1485). CHATHAM Ist.: Sappho Cove, common on sand beaches (no. 1486). INDEFATIGABLE IsL.: southeast side, common on sand dunes (no. 1487). James Ist.: northeast side, on sand_ beaches. SEYMOUR IsL., NorTH: Snodgrass and Heller. Further distr. S: U.S. We ind; S: Am/China. Trianthema L. T. Portulacastrum L. Sp. Pl. 223 (1753); Rob. (1), 144.— ALBEMARLE Ist.: Turtle Cove, fairly abundant on sand beaches (no. 1488); Villamil, in dry sandy soil in open places near sea level (no. 1489). Barrincton Ist.: Snodgrass and Vor. I] STEWART—BO'ANY OF THE GALAPAGOS ISLANDS 65 Heller. CHAR LES IsL.: common near the shore and in loose ashy soil at 450 ft. (nos. 1490-1491). CuatHam Isi.: An- dersson. DuNCAN Ist.: (no. 1492). GarDNER IsL. (near Hoop Ist.) : abundant in loose soil mixed with fragments of lava (no. 1493). Hoop Ist.: Snodgrass and Heller. INvDE- FATIGABLE Isu.: Andersson. JAMES Ist.: Andersson. SEy- MOUR Ibs., NoRTH and soutH: Snodgrass and Heller. Widely distributed on tropical shores. Aizoacea (?) sp. Sterile specimens of bushes 4-5 ft. high with succulent leaves were collected on BRATTLE Ist. and on the beach at Cormorant Bay, Cuartes Ist. The family is doubtful (nos. 1494-1495). PORTULACACEAE Portulaca L. P. oleracea L. Sp. Pl. 445 (1753); Rob. (1), 145.— Axincpon Ist.: common among rocks on the lower parts of the island (no. 1496). AtBemarLeE Isv.: Iguana Cove, abundant on the sides of the cliffs above the cove (no. 1498) ; Tagus Cove, common in tufaceous soil on the lower parts; Villamil, abundant in open places in the lower parts (no. 1497). Cuar.es Ist.: common around 1750 ft. during the dry sea- son; during the rainy season it occurs abundantly all over the lower parts of the island, (no. 1499). CuatuHam Isi.: Wreck Bay, common at 450 ft. (nos. 1500-1501). GarpNer Isv. (near Hoop Ist.) : Snodgrass and Heller. Hoop Ist.: occa- sional at 250 ft. (no. 1502). NargorouGH Isx.: north side, common in crevices in the lava (no. 1503). The fact that this species is found on such unfrequented islands as Abingdon and Narborough would seem to indicate that it might not have been distributed by intercommunication among the islands as suggested by Robinson, |. c. At the time Dr. Robinson’s paper was written it had only been found on the more frequented islands. Widely distributed. P. sp. (?). Sterile specimens of a species of Portulaca (?), were found on ABINGDON, BARRINGTON, BRATTLE, CHARLES, Jervis, and WENMAN Istanps. It is the P. sp.?, mentioned by Robinson, |. c. 145. (nos. 1504-1510). 66 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. P. sp— CHATHAM IsL.: Sappho Cove, low bushes on lava in the vicinity of the coast. The specimen has a brownish-gray succulent stem, linear leaves, and rather large pinkish-white flowers. It is a new species to the islands and possibly to science, but the specimen is too poor for accurate description (moms 1): CARYOPHYLLACEAE Drymaria Willd. D. cordata (L.) Willd. ex Roem. & Sch. Syst. V. 406 (1819). Holosteum cordatum L. Sp. Pl. 88 (1753). Dry- maria cordata Willd. 1. c.; Rob. (1), 145.—ALBEMARLE ISL. : Tagus Cove, abundant in open places on the inner wall of the crater at 4000 ft. (no. 1512); Villamil; common in moist places, 600-1300 ft., (no. 1513). Cartes IsL.: occasional in vegetable mold among rocks, 1000-1450 ft., (nos. 1514- 1516). CuHatHam Ist.: Wreck Bay, common above 900 ft. (no. 1517). James Ist.: Darwin. Widely distributed. MENISPERMACEAE Cissampelos L. C. galapagensis nov. sp. Scandens lignosa, caulibus canaliculatis glabris subtus glaucis; folio- rum laminis peltatis triangularibus vel subcordatis 4.3 cm. longis 4.7 cm. latis apice obtusis vel rotundatis mucronatis utrinque subglaucis, petiolis 8-44 mm. longis canaliculatis; inflorescentia mascula axillari cymosa longipedunculate ad pedunculi basis bractea membranacea praedita; sepalis orbiculari-rhombeis 1.5 mm. longis, nervo medio prominulo; corolla disciformi 1.2 mm. lata. INDEFATIGABLE IsL.: Academy Bay, abundant on bare lava in rather open woods near the coast. A species which is closely related to C. glaberrima St. Hil. but differs principally in hav- ing the male flowers in cymes instead of panicles, and in the sepals being orbicular rhombic instead of lanceolated, with a medium rib on each, (nos. 1518-1519). Plate III, figs. 9-10. Endemic. C. Pareira L. Sp. Pl. 1031 (1753) ; Rob. (1), 146—AsiNncG- pon Ist.: common above 500 ft. (no. 1523). ALBEMARLE Ist.: Cowley Bay, occasional at 1600-2000 ft., abundant on trees above 2000 ft.; Iguana Cove, common on trees and Vor, I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 67 bushes everywhere (no. 1522) ; Tagus Cove, common at 2000 ft.; Villamil, abundant covering rocks in a moist area on the lower parts some distance back from the shore, abundant throughout the transition and moist regions, (nos. 1520-1521 ve CHARLES Ist.: abundant in woodland at 1000 ft., covering rocks and trees at 1450 ft., (nos. 1524-1526). CHATHAM Ist.: Basso Point, abundant in woodland above 900 ft.: Wreck Bay, common throughout the wooded areas below 1000 ft. (no. 1527). Duncan Ist.: occasional on bushes at 1200 ft. (no. 1528). INpDEFATIGABLE IsL.: Academy Bay, abund- ant on trees above 100 ft., around 600 ft. it covers the trees and bushes with a dense growth, (nos. 1530-1531) ; northeast side, common above 300 ft.; southeast side, fairly common on bushes at 600 ft. (no. 1529). James Ist.: James Bay, abund- ant on trees and bushes above 1000 ft. (no. 1532); northeast side, fairly common above 400 ft. Nareoroucu Ist.: north side, (no. 1533). This species shows much variation in the size, shape, and amount of pubescence on the leaves. Further distr. general in tropical regions. ANONACEAE Anona L. A. cherimolia Mill. Gard. Dict. ed. VIII. n. 5 (768) Cuartes Isi.: forming a small grove at 1000 ft. Probably introduced, (no. 1535). Further distr. Mex., W. Ind., S. Am. A. glabra L. Sp. Pl. 537 (1753).—ALBEMARLE Isv.: Villa- mil, bushes and small trees in low moist places in the vicinity of the shore (no. 1536). Further distr. S. U. S., W. Ind. CRUCIFERAE Brassica L. B. arvensis (L.) Kze. Rev. Gen. I. 19 (1891). Sinapis arvensis L. Sp. Pl. 668 (1753). B. Sinapistrum Boiss. Voy. Esp. IT. 39 (1839-1845); Rob. (1), 146.—Cuartes Ist.: Andersson. Widely distributed. B. campestris L. Sp. Pl. 666 (1753); Rob. (1), 146.— Cuartes Ist.: Snodgrass and Heller. Cuatuam Ist.: Wreck 68 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. Bay, around habitations at 900 ft. Probably introduced. Widely distributed. Coronopus Ludw. C. didymus (L.) Sm. FI. Brit. Il. 691 (1800). Lepidium didymum L. Mant. 92 (1767). Senebiera pinnatifida DC. Mém. Soc. Hist. Nat. Par. VII. 144, t. 9 (1799); Rob. (1), 146.—ALBEMARLE IsL.: Villamil, abundant in open grassy country around 1300 ft. (no. 1540). James Isz.: Darwin. Widely distributed. Lepidium L. L. virginicum L. Sp. Pl. 645 (1753).—CuHatuHam Ist: Wreck Bay, around habitations, probably introduced, (no. 1538). Widely distributed. ee i. R. sativus L. Sp. Pl. ed. 2, 935 (1763); Rob. (1), 146.— CuHaARLEs Isut.: Andersson. Pore AM Iszt.: Wreck Bay, in cultivated ground. Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 81 FATIGABLE Ist.: Andersson. James Ist.: Darwin. Nar- porouGH IsL.: north side, common on lava beds. SEYMOUR Ist., souTH: Snodgrass and Heller. Widely distributed. Var. anacanthus Rob. (1), 157.—ALBEMARLE IsL.: Tagus Cove, common in tufaceous soil on the tops and sides of the hills surrounding the cove (no. 1730). Endemic. T. sericeus Anderss. (1), 245, (2), 107; Rob. (1), 157.— Cuartes Ist.: occasional among rocks along the shore (no. 1732). Cuatuam Ist.: Andersson. Endemic. T. sp. Rob. (1), 157.—CuLpepper Isv.: sterile specimens, evidently of the same species as those collected by Snodgrass and Heller at this place, were found by F. X. Williams, (no. ASI) e RUTACEAE Zanthoxylum L. Z. Fagara (L.) Sarg. Gard. & For. III. 186 (1890). Schin- us Fagara L. Sp. Pl. 389 (1753). Z. Pterota HBK. Nov. Gen. & Sp. VI. 3 (1823); Rob. (1), 158.—ABinGpoN Tisrs: common bushes above 450 ft.; above 1000 ft., small trees which are much covered with epiphytes. In the region around 1650 ft. they are scattered and somewhat stunted in appear- ance, (no. 1733). ALBEMARLE IsL.: Cowley Bay, occasional bushes at 600 ft., larger and more abundant above 1000 ft., (no. 1734); Iguana Cove, common bushes, forming dense thickets in places, (no. 1735); Tagus Cove, common bushes, 300-2200 ft.: Villamil, bushes on lava near the coast, increas- ing in size with the elevation until they form small forest trees around 1300 ft.; above 1500 ft. they form bushes or low stunted trees. A few specimens were found on the rim of the crater at 3150 ft., and on the floor at 2750 ft., (no. 1737). Cartes Ist.; common bushes on the lower parts, small trees around 1000 ft., very abundant on the leeward sides of most of the craters 1000-1450 ft., (nos. 1738-1739). CHATHAM Ist.: Wreck Bay, common bushes and small trees, 150-800 ft., (no. 1740). Duncan IsL.: common bushes above 900 ft.; around 1200 ft. it forms low trees, (no. 1741). Hoop Ist.: low trees in a very restricted area around 600 ft. where it forms a belt around the top of the island (no. 1742). INDE- FATIGABLE Is~.: Academy Bay, bushes in the vicinity of the 82 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H SER. shore, forest trees 20-40 ft. high above 350 ft., (no. 1744) ; northwest side, common bushes above 100 ft., trees above 700 ft., (no. 1743); southeast side, common bushes, forming almost impenetrable thickets, above 450 ft. It does not grow as large here as it does at Academy Bay. JAmes IsL.: James Bay, common bushes on the lower parts, small forest trees around 2000 ft., stunted bushes around 2850 ft. where it is exposed to the wind; northeast side, common bushes above 350 ft. (nos. 1745-1746). NarporoucH Ist.: south side, Snod- grass and Heller. This species seems to be one of the favorite host plants for Phoradendron Henslovii. Owing to its long recurved thorns it is one of the most disagreeable bushes to contend with when traveling on the lower parts of the islands. Further distr. S. U. S., Mex., W. Ind., S. Am. SIMARUBACEAE Castela Turp. C. galapageia Hook. f. (3), 229, (4), 262; Rob. (1), 158.— ALBEMARLE Ist.: Cowley Bay, low bushes to 1100 ft. CuHatHaAM Ist.: Darwin; Baur. Hoop Ist.: low bushes around 600 ft.; no specimens were taken. Endemic. Forma albemarlensis Rob. (1), 158. Forma jervensis Rob. (1), 159.—Arsemart-e Ist.: Tagus Cove, common bushes on the lower parts (no. 1747) ; Villamil, common bushes on lava beds to 200 ft. (no. 1762). InpDEFATIGABLE IsL.: northeast side, common bushes 6-8 ft. high in loose ashy soil near the shore. Stem unarmed; leaves for the most part cuneate with revolute margins, but some are obtusely oblong and mucronate as in the specimens from Albemarle, (no. 1748); northeast side, occasional bushes on the lower parts. The specimens from this part of the island are armed, leaves usually oblong obtuse mucronate, but some are lance-oblong acute, (no. 1749) ; southeast side, common bushes to 600 ft. Stem un- armed; leaves oblong obtuse mucronate, (no. 1750). JERvIs Ist.: Baur. Considering the great variability of the forms as shown by subsequent specimens, the form jervensis seems to agree rather too closely with the type specimen of form albe- marlensis to be considered as a good form. NARBOROUGH Ist.: north side, bushes 5-6 ft. high on lava beds (no. 1651). Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 83 Forma bindloensis Rob. (1), 158.—BrnpbLoe Ist. : common bushes. The specimens have the stem armed and many of the leaves are obtuse cuneate, (no. 1752). Forma carolensis Rob. (1), 158.—Axrnepon IsL.: common bushes to 500 ft. (no. 1753). Cartes Ist.: bushes 6-7 ft. high to 700 ft. Specimens taken below 350 ft. have larger leaves than do those from around 700 ft., (nos. 1758-1759). CuatHam Ist.: Wreck Bay, common bushes on the lower parts. The type specimen of the species was collected on this island by Darwin and is described by Hook. f., 1. c., as being unarmed with the leaves linear lanceolate acute. The specimen under consideration has the stem armed with the leaves varying from oblong obtuse to spatulate. It resembles the form caro- lensis very much, (no. 1757). SryMmour IsL., souTH: occa- sional bushes. Stem unarmed, leaves similar to those described by Rob. |. c., (no. 1760). Forma duncanensis Rob. (1), 159.—Barrincton IsL.: bushes with procumbent armed branches, leaves oblanceolate acute with revolute margins, .4-1 cm. long, (no. 1754). Dun- cAN Ist.: prostrate bushes above 300 ft. The specimen is armed with very strong spines, leaves oblanceolate with mar- gins strongly revolute, .9-1.6 cm. long. The type specimen is evidently a young branch, the leaves at the base of which tend to assume the revolute form. There is a single weak spine on the type specimen, (no. 1755). Jervis IsL.: occasional pros- trate bushes to 1050 ft. Stem armed, leaves attenuate obtuse, .6-1.9 cm. long, (no. 1756). There is much variation in the arming of the stems and in the size of the leaves in the speci- mens from the different islands, as well as in specimens from the same island. The specimen from Barrington has the larg- est leaves and spines intermediate in size, that from Duncan has the leaves intermediate in size and the largest spines, while the specimen from Jervis has the smallest spines. The most important character which the specimens from the different islands have in common is the procumbent habit. Forma jacobensis Rob. (1), 159.—James Ist.: James Bay, bushes 4-5 ft. high, fairly common below 300 ft. Stem armed, leaves broadly oblong obtuse to lance-oblong acute, with or without revolute margins on the same specimen, (no. 1761). 84 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47TH Ser. From the above it can be seen that if formal differences occur in this species, such differences are not confined to a single island, as it often happens that specimens from different parts of the same island show quite as marked variations as do speci- mens from different islands. BURSERACEAE Bursera L. B. graveolens (HBK.) Trian. & Planch. Ann. Sci. Nat. 5, XIV. 303 (1872). Elaphrium graveolens HBK. Nov. Gen. & Sp. VII. 31 (1825). B. graveolens Trian. & Planch. |. c.; Rob. (1), 159.—Asinepon IsLt.: common trees to 1000 ft., below 400 ft. they are small and scattered, (no. 1762). ALBE- MARLE Ist.: Banks Bay, common trees to 1700 ft., according to F. X. Williams; Cowley Bay, small trees above 400 ft., com- mon trees, 3-4 inches in diameter and 12-15 ft. high, around 1200 ft., large spreading trees much infested with Usnea lon- gissima above 2000 ft.; Elizabeth Bay, Snodgrass and Heller; Iguana Cove, occasional small trees to 400 ft. The small size and scarcity of this species here may be due to the more moist conditions which prevail, (no. 1765); Tagus Cove, common trees in tufaceous soil on the lower parts and on the sides of the mountain to 2000 ft.; Villamil, low spreading trees com- mon to 550 ft. BArRriINGTon IsL.: small trees, leafless in Octo- ber and July, much infested with Roccela peruensis. BINDLOE IsL.: northeast side, common trees in tufaceous soil. CHARLES IsL.: common trees to 1000 ft. CHATHAM IsL.: Basso Point, common trees to above 1000 ft.; Sappho Cove, common trees to above 800 ft.; Wreck Bay, common trees to 700 ft. CuL- PEPPER IsL.: low spreading trees, apparently of this species, were seen around the top of the island. GARDNER IsL. (near Hoop Ist.) : small trees all over the island (no. 1767). Hoop Ist.: trees 12-18 ft. high, common on all sides of the island except the south, where they seem to be rather scarce, (no. 1768). INDEFATIGABLE IsL.: Academy Bay, common trees to 350 ft. (no. 1769) ; north side, small trees and bushes on lava beds; northwest side, common trees to 750 ft., attaining their largest size around 600 ft.; southeast side, common trees below 500 ft. JAmes Ist.: James Bay, abundant below 1000 ft. ; Vor. 1) STEWART—BOTANY OF THE GALAPAGOS ISLANDS 85 north side, extending to above 1800 ft. according to F. X. Williams. JeRvis Ist.: small trees on the lower parts (no. 1771). NarporouGu Ist.: north side, small trees on lava beds (no. 1772). Tower Ist.: small trees, much infested with lichens. This species forms one of the most common trees in the dry and transition regions on the islands where it occurs. It seldom attains a great height, usually having a broadly spreading crown and a short thick trunk. Its absence from Duncan Island is rather peculiar, as it is found on all of the adjacent islands, and the conditions here do not seem to be such as would inhibit its growth. Further distr. Mex., W. Ind., S. Am. to Peru. B. malacophylla Rob. (1), 160.—SryMour Ins., NortH (?) and soutH: Snodgrass and Heller. At both the times our party visited south Seymour, viz. in July and November, the Bursera trees were out of foliage. So far as is known this species does not occur on the north side of Indefatigable although this island is separated from Seymour by a channel which is only about a half mile in width and is probably of comparatively recent origin. Endemic. POLYGALACEAE Polygala L. P. Anderssonii Rob. (1), 160. P. puberula Anderss. (1), 232, (2), 100.—INDEFATIGABLE IsL.: northwest side, Anders- son; Baur. Endemic. P. galapageia Hook. f. (3), 233; Rob. (1), 160.—Asinc- pon Ist.: fairly abundant on the lava beds on the lower parts (no. 1773). ALBEMARLE IsL.: Cowley Bay, not abundant (no. 1775) ; Tagus Cove, abundant from the beach to 600 ft. (no. 1774). Brnpiok Ist.: occasional in tufaceous soil near the shore (no. 1776). CuHarwes Ist.: Darwin; Andersson; Baur. Cuatuam Ist.: Wreck Bay, abundant in sandy soil near the shore (no. 1777); north side, Baur. INDEFATIGABLE IsL.: north side, abundant in light ashy soil near the shore (no. 1778) ; northwest side, abundant in tufaceous soil near the shore (no. 1779). James IsL.: northeast side, specimens seen at 200 ft. Jervis Isxt.: Baur. Endemic. 86 CALIFORNIA ACADEMY OF SCIENCES (Proc. 47H Ser. Var. insularis Rob. (1), 161. P. obovata Hook. f. (3), 233. —ALBEMARLE IsL.: Macrae. CHarves IsL.: Cormorant Bay, abundant on sand beaches (no. 1780). CHatHam Is_.: Sappho Cove, abundant on sand beaches (no. 1782). INDE- FATIGABLE Ist.: Academy Bay, common on the lower parts (no. 1783). James Isx.: northeast side. Jervis Ist.: Baur. Endemic. EUPHORBIACEAE Acalypha L. A. Adamsii Rob. (1), 161—Cuatruam Ist.: Wreck Bay, Baur. Endemic. A. albemarlensis Rob. (1), 163—ALBemarte Ist.: Iguana Cove, occasional among dense vegetation at 300 ft. (no. 1784) ; Tagus Cove, Snodgrass and Heller. Endemic. A. Baurii Rob. & Greenm. (1), 144, 148; Rob. (1), 163.— ALBEMARLE Ist.: Villamil, common in open woodland at 1300 ft. (no. 1793). CHatTHam Ist.: Wreck Bay, Baur. En- demic. A. chathamensis Rob. (1), 163——CuHaAtHam Ist.: Basso Point, occasional among rocks at 800 ft. (no. 1785) ; Wreck Bay, Snodgrass and Heller. Endemic. A. cordifolia Hook. f. (3), 186; Rob. (1), 163.—CHARLES Ist.: Darwin. CHATHAM IsL.: Andersson. Identity doubt- ful acc. to Rob. 1. c. Endemic. A. diffusa Anderss. (1), 240, (2), 104, t. 14, f. 4; Rob. (1), 163.—ALBEMARLE IsL.: Cowley Bay, Andersson. CHATHAM Ist.: Wreck Bay, A. Agassiz. Endemic. A. flaccida Hook. f. (3), 186; Rob. (1), 164.—James Ist. : Darwin. Endemic. A. parvula Hook. f. (3), 185; Rob. (1), 164.—ALBEMARLE Ist.: Iguana Cove, Snodgrass and Heller; Tagus Cove, com- mon ‘above 1500 ft. (no. 1786); Villamil, Baur. CHARLES IsL.: common in rather open brushy country around 1,100 ft. (no. 1787). Endemic. A. reniformis Hook. f. (3), 187; Rob. (1), 164.—CHARLEs Ist.: Darwin. Endemic. Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 87 A. sericea Anderss. (1), 238, (2), 103, t. 14, f. 1; Rob. (1), 164.—ABsinopon IsL.: occasional to 500 ft., abundant above this elevation. The specimens from the different elevations are similar in the size of the leaves and in the pubescence, (nos. 1788-1791). ALBEMARLE Ist.: Andersson. BINDLOE IsL.: Baur; Snodgrass and Heller. Endemic. A. spicata Anderss. (1), 239, (2), 104, t. 14, f. 3; Rob. (1), 164.—ALBEMARLE Ist.: Iguana Cove, abundant on the sides of the cliff above the cove (no. 1792). CHaArteEs IsL.: occa- sional at sea level and at 1200 ft. (nos. 1794-1795). CHATHAM Isu.: north side, Andersson. DuNcAN Ist.: Baur. GARDNER Isx. (near Hoop Isu.): (no. 1796). Hoop Isi.: rare around 300 ft. (no. 1797). INpDEFATIGABLE IsL.: Academy Bay, occasional below 75 ft. (no. 1800) ; northwest side, occasional in tufaceous soil near the shore (no. 1799). Jervis Ist.: Baur. Endemic. A. strobilifera Hook. f. (3), 187; Rob. (1), 164.—ALBE- MARLE Ist.: Cowley Bay, Baur. CHATHAM Ist.: north side, Darwin; Andersson; Baur. Endemic. A. velutina Hook. f. (3), 186; Rob. (1), 164—CHaRLEs Ist.: Darwin; Andersson; Baur. CHATHAM Ist.: Wreck Bay, common in open shady woods around 700 ft. (no. 1801). Endemic. Var. minor Hook. f. (3), 187; Rob. (1), 165.—CHaARLEs Ist.: Darwin; Baur. Endemic. A. sp.—ALBEMARLE IsL.: Cowley Bay, occasional below 300 ft. Sterile and indeterminate, (no. 1802). A. sp.—INDEFATIGABLE Ist.: north side, common at 250 ft. Indeterminate. Both of the above specimens probably be- long to species already described from the islands. A. sp. Rob. (1), 165. A. parvula var. cordifolia ? Rob. & Greenm. (1), 148.—Barrineton Ist.: Baur. Endemic. A. sp. Rob. (1), 165.—Barrineton Isi.: Snodgrass and Heller. Endemic. A. sp. Rob. (1), 165.—BarrineTon Ist.: Baur. Endemic. A. sp. Rob. (1), 165.—INpDEFATIGABLE IsL.: south of Con- way Bay, Baur. Endemic. January 12, 1911 88 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. A. sp. Rob. (1), 165.—A. parvula var. flaccida Rob. & Greenm. (1), 148.—Duwncan IsL.: Baur. Croton L. C. Scouleri Hook. f. (3), 188; Rob. (1), 165.—ALBEMARLE Ist.: Villamil, common bushes, 100-350 ft., (no. 1804). BarRINGTON Ist.: bushes 6-8 ft. high all over the island (no. 1805). BrinpLoe Isi.: common bushes in tufaceous soil (no. 1806). Bratrve Isi.: low bushes, nearly leafless in October, (no. 1807). Cartes Ist.: Snodgrass and Heller, approach- ing var. incanus according to Rob. 1. c. CHatuam Isi.: north side, Darwin; Baur. Hoop Ist.: bushes 10 ft. and more in height all over the island (no. 1808). INDEFATIGABLE IsL.: Academy Bay, occasional bushes to 550 ft.; southeast side, common bushes all over the lower parts, (nos. 1809-1810). James Isi.: Douglas; Scouler; Andersson; James Bay, Snod- grass and Heller. Jervis Ist.: bushes 4-5 ft. high all over the island (nos. 1812-1814). NarsorouGu Ist.: south side, Snodgrass and Heller. Tower Ist.: Baur. Endemic. Var. albescens Muell. Arg. in DC. Prodr. XV. pt. 2, 605 (1862) ; Rob. (1), 165 —ALBEMARLE Ist.: Andersson; Eliz- abeth Bay, Snodgrass and Heller; Tagus Cove, occasional bushes to 4000 ft. (no. 1816). BrnpLoe IsL.: Baur. CHARLES Ist.: Andersson; A. Agassiz. CHATHAM IsL.: Basso Point, common bushes to above 900 ft. (no. 1819); Wreck Bay, small trees and bushes on the lower parts (nos. 1817-1818). INDEFATIGABLE IsL.: north side, bushes 6-7 ft. high at 300 ft. (no. 1820). James Ist.: Andersson; northeast side, small bushes on lava beds (no. 1821). Endemic. Forma microphyllus Muell. Arg. 1. c.; Rob. (1), 166.— ALBEMARLE Isu.: Andersson. Endemic. Var. brevifolius Muell. Arg. !. c. C. brevifolius Anderss. (1), 241, (2), 105. Var. brevifolius Muell. Arg. 1. c.; Rob. (1), 166.—ABrincpon IsL.: common bushes 4-5 ft. high below 1000 ft. (no. 1822). ALBEMARLE IsL.: Iguana Cove, Snod- grass and Heller. BinDLOoE Ist.: common bushes (no. 1832). Cuar_es Ist.: bushes rather characteristic of the region be- tween 650 and 1100 ft. This species becomes more abundant with the increase in elevation and forms a belt, around the base Vor. I] STEWART—BOTANY OF THE GALAPAGOS ISLANDS 89 of the central mountain, which is noticeable from Black Beach Road during the dry season when most of the other vegetation is leafless or has the leaves very much reduced, (nos. 1823- 1824). CuLPEpPPER Ist.: F. X. Williams, collector. Croton bushes appear to be very abundant on the top of the island. GARDNER IsL. (near Hoop Ist.) : Snodgrass and Heller. Hoop IsL.: common bushes (no. 1826). INDEFATIGABLE ISL.: Academy Bay, occasional bushes and small trees 10-15 ft. high to 300 ft. (nos. 1828-1829) ; northwest side, bushes 6-10 ft. high (no. 1827). James Ist.: James Bay, common bushes to 1000 ft. (no. 1830). Srymour Ist., NortH: Snodgrass and Heller. \WENMAN Ist.: slender trees and bushes (no. 1833). Endemic. Var. glabriusculus nov. var. Foliis ovatis denticulatis acutis utrinque sparsim pubescentibus; pilis aliis simplicibus aliis stellatis lamina circa 4 cm. longa 2.4 cm. lata. ABINGDON IsL.: small trees and bushes, 1000-1650 ft., (no. 1834). Csi eresi elects a is die 6 esse ws sere saprorpoddjod aioe seane eceiues ire) ae. fe 58 9:18 (8/608),0' 88) ese umosorayd urounq | a[syeig | copurg | seddading | wemen) | uo}3uLueg | apemoqry | uopsuiqy | soBedeyen | ysnosoqieny a1qesizejopuy ‘PH Jeupiey “YO soupiey panuyuoy SANVIST SODVdVIV) AHL NOdN SALAHAOLVWAAIS GNV SALAHAOGIMAL AO NOLLAGINISIG] AUL ONIMONHS W1dV I, STEWART—BOTANY OF THE GALAPAGOS ISLANDS 165 Vor. I] seddading ysnoroqieny a1qesiyejopuy “PH Joupiey “YO Joupren +++ pen Pee are Reva otatere ucrsvoraKousvuekelery chePera arene Susan Ec Oar OM OR ecio tad CeO oD tc vs[naip eprsuy steirhsl cea iopellctaeleveke erp Ts eqyipoiydeurioy eroydoyyuy Oe BER Be as elueusie eprydouwuy AVANINVU) eIOJI}L] “IVA eulieu seleN aVaOVaVIVN Si acherNecerneteneteenrs Wor oroteretetarare eunjuem viddny rey snyeurjood uojasoulejog AVAOVNOLASOWVLOg umu0,oYoIp Buavarscona ehavese lovarere ev auatsllovensys stares umnyeurdu0s TL ROSACEA tunzearyo wntpodooé’] AVAOVIGOdODAT ayiewmeqy uopaurqy sosedeyer SANVIST SODVdV1V‘) panuyuoy) SHL NOdN SALAHdOLVWYAdS AGNV SALAHdOGINaLG AO NOILNATALSIC] AML ONIMOHS ATV [ [Proc. 4TH Ser. CALIFORNIA ACADEMY OF SCIENCES 166 STi [Sams = ai aie male nmaay cohu(mevi ander imaen sine cts ncrinnprecel> ciscoirigte eases eecenroge esoyid ine fay is eis ier i Ti fart er 7 ee ee oe eAyoryseSaur Ef a zn 4 = ahs ee ee ec - ster re ee ee ce sisuatyeq srysor8ery SWS IPN SEE WRT leierk ala anne aiereiorr eva lensios eng afro tae ap poIpur oursnopa + et aie.d>4'ja)'e @0.6)0).07 6: (6) 6149.6 04:4) 00-816) sljeuinsurs BLIepSIG ee} ie se tat: Thee eT TAs [fF | Pam | abst nakeneschare%eke ec edeseceks uniydAsov umtuszoAjyorq tet fd ft ese ee tae nik a ntide s toa cgrm a er dinigarte ey eyeipes i 7 ee cn sueSaya iL a7 ee ae (ee (ee (coo Coat epodostur sts0[yO ae PN AIP Too AI avaiecnsiaree aiain nugimierncnaieicios, ties ginieis i biele ds =f aie ae ae -+- + + + + + + 4+ Ole 606 2-08 65a 8 ee (0c mee 6) 8 618, 0.6 0 0,0c8 snyyuroe syed mpilice pay ee ee suepnues3 ae es afi, fh a ipeerbistherens teneieze ee cetenayetient snyAydoyorjstp sniyouag cr] (TEs ree es _ ee ee ee eee esopid wnojaynog - tl fot Pd ff | peccecscectcceesesetet esses eens PSOTIIA reg (re re erg ray fe ode ae] geld ete tee bene cee the See ee ae eyrordsqns ES z| |E|? i i i: panuyuoy) SANVIS[T SODVdVIV5) JHL NOdN SALAHdOLVWAAdS GNV SALAHdOGIWALd AO NOILNAIMLSIGQ] JHL ONIMOHS ATV | STEWART—BOTANY OF THE GALAPAGOS ISLANDS 167 Vor. I] Ty a eel ae Se (ae eae oe Resear ooananaar as a Fees (Sra Lad fee ors en) letersi a en Ne eC umurjosas ae ee ae edliah pM otecle ley i|ivew to scemteerairetatarerstoinve saneratpia umurpnonpnur mo Kaa (eM He! Viet (Ode a (2 IP en eee eer eee rene ee Tee aqjour =e et fe fea |operates cri snutu ‘rea oe fered er alee Be Po MMe [teeta scsi aie a os sete a ene ts umnpneonaty fi Te cfg Hess era Ue i art fe hel Berne aecce prime ee umyeuruas eae epee 27 (ee nn enn umny_NoOsey med I | a CLT ica taal ha ae cal |p Mh by sat ome nn ee ane ee pra ene eaeres sue dere Cole wimnuoj[ood UWInoOUv = ot 0) d, 6-0) 6 eye tote; scale vere) s eve eee, 8 ole sniieyas snuowsidg oT PT (I ae Nee (oe WR logo ene ct Ab coneoce ncn rngonorac eye8r1a ari re Geom fh I |e) biavave-ave crorn cuss slanel slats syetelsteoaie vbre vyeuoronu i na ae Bp] frcee cece eee sees eee e eee e ees puvAo[pury as e715 Nee font || is (ei ae tel cn Fa PociimscdantanmnoaennGaccadco de dp sturzosTy + + Ra nas ee a a Ra edn eh ge SISUs] IP Wed |e evoTyooyda’T Set Te UP oI ih lt elk cmeen erence tte ecseecaen Staccees Camas eqyeqound SpA Ei cafe G2 re NL MY ese teccine aves alsonse vise qhonetewwtdeenhsee eAYoRIsIp voyooug PIPISIEISIEIELE RIE Eee Zizi e Plz le ESE Yel ehad iatediel b tole (ite Petia Pisteonael We = pau 1B = yea Vet ten ce T= Paka =n t=) Va =a ak ob MLE E; (EVE VS TELE IF Ie Bele a 3 5 = 9 g Sle a panuizuo) SGNWIST SOOVdVIV‘) AHL NOdn SHULAHdOLVWUYAdS GNV SHLAHdOdIYaALG AO NOILNATYLSIGQ]. HHL ONIMOHS ATV] January 13, 1911 [Proc. 4TH Ser. CALIFORNIA ACADEMY OF SCIENCES 168 + ear + ale]? | 2 e)2/3|8 BIS leis 5 R |e a ing aqqesyyeyepuy “PH Joupien "yO soupreg saddadmg ++++ fb weyqeyy anqerg | SOTpUrg | uoqsuLeg | ie > SoM Per sabe ae ++ apeweqry | uopauiqy | sosedeyen | ricer ER Oat oe eS eqerysor edig Bsbahe Sh ciachsis! Stu a palare ete wnjepunoas wniydryouays sEaiatege 'satekove couavotenstage sueveasraterens seenatts snorursata yas Shes tos heer sernsenecaneeiehe pater ers ercteteetene snoiput Se 8 chi ee Pele hake a CRE sisuosuiwop snjoqorods aise: E slsnocereles okeisle,eiCuast/aveyesnete eis ater rece -ds 4) 4/6. G0. e/ al eMeierele ees 6: 6.0/6.6: \006 lop eieibile eherb kh 6.6 ‘ds Suey ahai's She..9),8.\9 lan a'fe, O101478p- Oe ase sues: axe winye[NoIqolos PRS BLS Oe Ee oe he Ce 6 fe ie Mieke. e ise cele uinye][iotuad aralive Le Kerelece)9 peng hel ayelealeholeierevele. olan oe wimnyo14sIp Pre ee eC Oe BR et PORCe a MON RT FC) wmnyesn{uoo Ce ee ee reece s eee esse sesese suaosouro unyedseg panuuo7y SANV1ST SODVdV1V‘) AHL NOdA SALAHdOLVWYAdS GNV SALAHdOGIMALG AO NOILAGIULSIGG AHL ONIMOHS ATaV STEWART—BOTANY OF THE GALAPAGOS ISLANDS 169 Vor. I) + snAyor4sty ah aleve ishsilecahe jel enetexelie ere eieia7s\isia\aleceleye stsueuitueins || ees eee no teed | eel | ad Mecemeteuercoderatolenevoueicsoietocen-bedshey=KecsKensmnante snso3tiys a at. ak weheie lea) enshevereteretets el sieusratai's snjnui09 ‘IPA ab 25 ab Svelavalevapana-s) eieliewsite) Sie) sho) auc, enelisibjis.easitd snsoursiqns mS | Se 1] ei] Recs \ceke er atkesel ote vse” eek sxelsta)inn =) srs)inifeystai size) aie snpunjo1 Sep eel swe: ofeguershshscells;elrsceme: ai sis, a. dqe\ielis snAyorysdjod ak She feleyeyle (oNerienie (one e/elleke) efseloa) velere, slvei- ere! eels HSTDY ailal alleys \elleeheeKetetehelelevede kellensle(aedece eteKevace s1epn3yy +\e+|+|+ ++ +++ ore lepe lee vee e.nre see foie ove es, cee 0 lene 00 snyestaory + + a 0) a eva cele ee ice (ole eee = wa ene af esas, snijoyipueis Se ee 2 Sp alee ed [tall Maienentevfaxexevterre ie soreretere.evaatexenstege sutama stsuosedeyes +++ @ ejrale leuerevellelele eke: ous lafsieveteketshefelerelevene snqzuetnosa @ (oles. 0/ 0/0 (,[e\el/etae) siete (ee wel or tsa 's « whee snyiojuoo ++ OCILCROO RC Dt et bth eCns eT ocaOn VCLCeE yt sfyovyshyorsq | t++H4 + B08) 0)6..4>0 eee) eee 6) 6].0) 5) = invelsyes 6.60) 018 snjeq4siie sniadAg AVAOVUAGAD uewus | IOMOL | ammoutéag | deddading weyqieqD uo0yBULLIE g ayreuraqiy | uopaurqy | sosedeyern | ysnosoqieny So toupresy a]qesyeyopuy ‘PH dJoupies panuyuoy SONVIS] SODVAVIV) AHL NOdA SALAHAOLVWAAMS GNV SALAHdOGIMALG AO NOILAATYISIG] AHL ONIMOHS ATAVL (Proc. 4TH Ser. CALIFORNIA ACADEMY OF SCIENCES 170 ysnoroqien aqqesyejopuy ‘PH Jouprey meqyeyD apeulaqry “GD souprey sosedeesg SUe|NSUT vISpUeTTLL AVAOVITHNOUNG sreperesenefelrsaVeNeu, MrtrecansanneNecRaseon Settee eine -ds aye she) alice Soeahrar aah arahar aos haysuc ays ees sour vuwe'y AVAOVNNA'T BCCI a Ot no EET pe aero: eyorayd enapog O70 OD et eer R OLD le te One Bia 8 eon Ere Eevee e,'e eruind esull[Ay Sea UecsSauatr Deeanasces Sn saaeoe eyjueior eydivormoyy sfevelatietshatizagyterenstsysqirehtuesth ty aekalecoicPho ere eyAydip PrieOL yy, Shu. lendte) gia qar a Maal alrase 0 “suede st[Aystiquity evel seal ctavers ssitieiMejntey sfersicesate ayer sv, taekeksee vyeqnur Bch elinl tren etetanay selehen tara slap sMepsraeyay tat Sater esoqnqsy Sie nteiasefajene Sinister aasaaeena ge eyeqideo steyooaysy ee ee eC ee ae ee er a ee ee ce ‘ds panuyuo7 SANVIST SOOVdVIV‘) AHL NOdN SALAHdOLVWUAIGS GNV SALAHdOUIMALG AO NOILNAINLSIG] HHL ONIMOUS ATV | STEWART—BOTANY OF THE GALAPAGOS ISLANDS 171 Vor. IJ | + +/+ + + ol stele teVeuehevey eter shes ot evegenenatent= die + + ales +|+ a + +/+ + += + + Sralictealla + et err Ste |icte SIPISIETEEEE le 2 ESE ElEVELEIE/e BLS le |e le e/BlAle ls lee lelelelele ls leis 5 5 & Fe fot |e ped teed [Mt tS atce| (ok dead sy a | > 3 B13 = Porgd|(Eeia yearby Ole ese ilicoun| (ea R 8 He 5 g jo]? 12 = s a|e oO ae | Wie panurjuoy eye[noel eASIYJUOT saprolepnoryn stsdouoy ds erydo[ng wuinqeaids umipuopidq HVAOVAIHOAG ‘ds sisueqno varloing AVAOVAGITIANVANY RGranoN Recon ds euurg AVAOVNNV’) AVaOVAlayT Sree curete BOOVUI[AWIUIOD “RIOYIpnu vulfaurUIOD AVAOVNITANWO) SGNVISJT SODVdVIVL) IHL NOdN SALAHdOLVWAAdS UNV SHLAHdOCdINALG AO NOILNALALSIG] AHL ONIMOHS ATV (Proc. 4TH Ser. CALIFORNIA ACADEMY OF SCIENCES 172 oz eee 7 ee (nnn saprojded A 8 HD |e Oca) Femme tT Pa 11 eel Be AT sl en | omic mira oti tien Denon tac eyAydosoru oe ee oa mneg voltg oe ee ee sIIqep viueqoueg ee ee cence suenyse vAImapy AVAOVOILYL) | +++ + du dhe als of ak re eye boa S%e whens chebe) ais) a) earn a“s,a 0,4: aetna © ri4yJeMayS + CCC ee ee ee ee . lssei8poug at Cee ee ee eer ee eso[nuies ee Oe a SR eT A AE atte wanseartetenel ae) atasne, aieletal shelielees:ake erates eyejorjod + Wrasse a) Oar ere So's o\ble vie! Sie) aie! alo ore,0°e%e,e, equitisnyqo pe ee meee reer reer sree ee tinesesece saploles + + + + ©) @) 60:)6. @, 4*6) @ fe, 01/0'\6. 0) 6:5) 6,046) 6 00 8.00 58 stsuasedrye3 =} Fe | (Pe ||) (et Wee! | | FL a acc ectCe Perio stuJojyesey eruosodeg AVaOVa Ad ueUlUa My JaMo], anoutf4ag tuqiedS ysnoioqieny aqqesizejopuy PH 7eupsed "AD Soupres) saddading uo Wu apemaqy uopsulqy sosedryes) panuyuo 7d SANVIS] SODVdWIV5) AHL NOdN SALAHAOLVWUAMS GNV SALAHAOGINALG JO NOILAAIUISIG. AHL ONIMOHS ATAV STEWART—BOTANY OF THE GALAPAGOS ISLANDS 173 Vor. I] 3 episis ate ab ate Sioa, afer eeele ates ener aver see lesehe eyVoIpel eloYyyUPUIEyLY AVAOVINVAVAY Pe TP A A ee ee ie WlstoncossocenocudooocondoDoodoc ds eBIUsOoITeS St | a | oe | PM cee et | eM | Hm Vt be He lh ares TNT I Meme Where avakerseee,seecsca raven eveis tere ontrenatewyoier Ghcreoemers ds + + 0108) eae (a ai70 fa ana/,aieXevlaleieiteie elie ferle, ‘ove lece lola lols ds xo[duyy AVAOVIGOdONAH) an Sel HI I incisiedetere)eietexs erehe ede ==, sei oitet siieiehleliotey 3 suasedvyes [once umzeurumnoe Steet I es SU, See Ni Tl] etetteiereieteiesereronevaveratsyolsvot hes elheistefs aI0e wunuo34jog AVAOVNODATOD + Oe Soke! 6.0 ey eiete. ob etelp\ ele! ofeleloleteleleie/ eles uinyeuroun + a + + + + + + 076. 6\' 6) 60) 6'10''6'\0':0/0) 6! e006) 0. 0) eels elole,eneeieie HAO|SUOFy et A SB AE EY Sa Hivtevere (ec ereielesere) esersalscéievsht) ete veheretesie untesedre3 bene) Wh LY cle teat Wall Pall hc texanevsneter sterevscetefstavses etene umMueLoy uorpuepesoyg AVAOVHINVAOT + + + eevel/e' 6 0) e056 revere ey ele~e 4: anehevehe ete! BI[OJoRoo|e BIdIn ale lelfls is lElBiP lel eieielelsiele BEI As ie | |) BRE 1s a a We ee VB ie J }2]5 |. Bs S/F 8 hpelele/eielsle(a/e/B]/e18 18 18 Bp 2/48 & e@|lo|8 | lp | 8 | o 1] 9 p || 5 & |e 2 La peta |p aete died Wats a/R ]o]e & 8 mlo 8 Sia ereall eae o = a} > e panuyuoy SANVIST SODVdV1V‘) AHL NOdn SHLAHAOLVWUAdS UNV SALAHdOdINALd 10 NOILNAGINLSIG] AHL ONIMOHS AV] [Proc. 47H Ser. CALIFORNIA ACADEMY OF SCIENCES 174 = 7 | | [ots QUO ISOM ouIsory e - rio: Sa ex pF ede sem; dpe bee s0 0/6. me she Blea yere,e @ alters: eliedoos eee) re |e ee (ee (fc (ernment syneorpnu ree) ble a (eis (woe 6/86 Bh BORIDES 8 Boe Bue Sera) \e BIISIUP] si oseresuetakeftatet lo PIETER IE IEIE IE el ela le le le SPI=l|EIElS iB leIZz/EIRIS |B la ls ct @ @ Kc] 8 ® 9 a]. g a ie ca aa: Pe ae |g ie g 3 2 A 5 o a panuyuo7y AHL NOdN SALAHAOLVWUAdS GNV SALAHdOdIWALd AO NOILNAINLSIGQ] AHL ONIMOHS ATAV [ STEWART—BOTANY OF THE GALAPAGOS ISLANDS 183 Vor. I] a ee te 3k au a =f + + + wlern 674 ens e%ee 6,0 ere elsnele rete 6.0 0) ole ef 6 4*¥ 650 eyeords fa SS ig ease ear err ee soe Be hogar cae (ees ha | ht aoe Re on ier cee Pt sueraes a elec Ralprctar saver tea niu canna eaaed biel meee al eerie all ec ecae Iae aeL eee nc oeee lance tate ean res apmoen = eral males iacrendetecas eae ec ee ner ss Se Faoeg el Sree | else - ly eles ae mae Se ee aonpios Pood el) Pdf seal | Gaarcsceracceacrocnicoctn: siavaaeenieno i lk i = Weererennae emer cae ange alls | Clewosereoncinench ema sausseeiqte fi Pee ee et ee eee ee HsWepy eydAeoy AVAOVIGUNOHANY de = len di [ole |eseoawaneectanmmeerang dcrepigin: “seh oft oe aft ae ak ab ail a6 eete ere oeleleleeele.ereverate-evecst evelehele, sie. e viasedryes fe Praia ter aCe er Ye Yar Sa Se ee TT tluossiopuy eyeshlog AVAIVIVOATOd eee Scoot il megt AM 2a leet c nn artotee ccs ce ace erence al a — 3 2/8 z A I a Pe 5 z = 9 § Sjer]F ia o 5 =. panuyuo7 SGNVIS[ SODVdV1V5) AHL NOdN SALAHdMOLVWUAdS GNV SHALAHAOUIWALG AO NOILNAIMLSIG: AHL DNIMOHS ATV January 13, 1911 [Proc. 4rH Ser. CALIFORNIA ACADEMY OF SCIENCES 184 ae are ae a Towle “seA it ze Sy lt ee coco snyojipues3 “ea + Cie acta ar Ma be be ede Pa at Ue Pd snynosniseqeys ‘IeA i i are ben bala es re bare ‘id fe ie snipoyraesq “ava co | | dah Wiens eames eh, snjAydosorm euioj a | a pray fa zi BNA, UND fie neretertie siete eictereirie “+ +suaosoqye “vA a ers (tp | i) Sp Se Tes tiats ati) Ml” Pll sci = a/asiatere eesereninsorerannier atconair luajnosg u0,019 =) dy (it (| [eon (Ve Wg PGi fo | (SEES 2 coe ceo a TLIC ds ee ee (eo ds del TP Pf Pater ciicieteieitinticin tte tecieccmas ceeeiires ds Be WM le TDs aS IG SN |i] sictraitia aiainn Sates necieG Girine gisinnsiereste ds te] | | ect re tees tec e erent eee eet eee ences -ds te] | | [ccc ee eect eee ence cece ceee teen eee ees ds tl | | [eter trte tee eee ects terete eter tree eens ds m7 es (Ae ie i cece JOUIUI “IRA BNE OSE AONE) SNP Seika: sconce acaiped Res non nd mete nan ner +++ purqnypaa ue ee ee vaoqtfiqosys Ae] /EVEVEIEIEIEIEIZIEIEEIZIEITIE ELE a18/8|8 BIPIFIBIS (SIS lee 1a] e Z| § a | 3 5 i e Be eel Meal Be WB ce hag = z| ||P : : panuuoy SGNVIS]T SODVdVIV5) AHL NOdN SALAHdOLVWAAdS UNV SHULAHAOdIUALY AO NOILNAINLSIC] AHL ONIMOHS ATV [ 185 GALAPAGOS ISLANDS STEWART—BOTANY OF THE Vor. I] aE cE rr eee OUTUITA + oe G5 exe 0.0 0.0 e-efele we) ee 0a" 6. 0u p06 ove eis, erlpojrum Ay} ple] [ecceece eee e essere ree eee ee ees nisuaAayg SC We A SE WN Upesesctstsre vtoces streets rtern ci stsrntereseas te Darneee + + | a ||| ee: eerscets ec cle. enedeneistare sisie (sieves elerstn eyeinjound = + + 00501 "e: Gi 6) 0.0) 0.0), 6,0, ole) ere 6. eee 10 wee eo bhe!-e. 5.5: eloyynyid + 4 a) siete yene).a08 Glee) oiee.ele le eeneiele, s eiqes “IPA A Aliph Mig Ul pd ebtisew.cies serne seein omsiolae sac euepnwumu Tall A Ee AN eee Ee TSE See alist he alo anticisensinocencniunter cups coort eonjorsou + CSO RCs OK ek Orn bat HOE) 0 i, Da SEO viasedryes alee feller rs Til afa| (ore fore ||| enone cuaenanen sonencisco sueyjoqep + Bee 0:16) 0.6), 0.6'.0 Te 10/¢ enele le Ciriani y Cet sTuUOjIyasInba re a ie sf Bi We |e oreepcontrenonan tn oo obonn. Ay esngip + Es fi ieee aaeeneercn er sIsua0]purq “3eA oh Be) | ele + Se Vip erie ancestries sola ctegeie ss eceetel eyeqnonze aE erabatey ate} ate Pale! shalatte) biases} akeleher el eretetarslate eqyenoide | +} + + oF = a) io ae Nea stage rom ee 2 syneorxe|due viqioydng SPIPZIPIF IEF elelelelelelFizlFielele Bele le le le lelelele leis lelk la lelalg le le Fda Pace H e BiB IBIS IBIeg |e /S/E1/R (B18 ra =. 5 5 Ls) val 5 ° 9 5 4 | 3 a ® | 3 §|/+/158 |o & e| |zle| |* Eales rc) a 2 panw4uo 7 SANWIST SODVdVIVL) AHL NOdN SALAHdOLVWAAdS GNV SALAHdOCINALG AO NOILNAINLISIC] AHL OSNIMOHS ATV (Proc. 4r Ser. CALIFORNIA ACADEMY OF SCIENCES 186 |} +++ ysnoloqieny alqesryejyopuy PH Jeupseyg seddadng meyyeqy 0).6/ 00 "s 0,070 00,0 eels 6 eB iele a anele 6 ee eee ds ayoyzy[eO AVAOVHOIMLITIVY 4b S10 (be <-OYe./ee Teva oy aira)6 fo Xe %0'10,(0)'d Iie lee) eUIsstyn qoyrue yy PiSi6se a ei el's oe save e608 ae ees eles ie 68 svoino edosjzef ab Bie Ce Oe Oe be ne Die ow Be Be 6 eyJoulouryy auvuoddryy CeCRCM AC eC CRO sVsheiieverisise}eis) seteliesecetereheracetuerey gs + CsKel aXe oyeseLe? eenlele; aie, 6 stsuouopsulqe “IPA ROU OPO Oe raL De stsuaasof vurz0} PUIG OROS ToDo stsueqoorl vu0j BUIIOJ BUIIOJ messy Teksatayeqaiarereraceveieiones sisuayosvo PUL} skedehehshosehacehstchesshhe sisusuo.sULIIeq eUIIO} _——————————————————————— panuyuoy SGNVIS]T SODVdV1V4) AHL NOdN SALAHAOLVWUAMS UNV SALAHAOUINALG JO NOILAGINLSI] AHL ONIMOHS aTavy, STEWART—BOTANY OF THE GALAPAGOS ISLANDS 187 Vor. I] Bp li | Gre cletras ee ec cine ee eqopiques eyoywNIy + é+ + Pr er et ee ee eee ee eee snjoqoyid snioyolog AVAOVITIL Se (ole Vee Waele de obese ieseqe sean oem PIOJIUIA STA at + =—- + © @), 0° @) ©).0)-ee- olele-vlielejsie euete:t)ecs-600, 0-40 \@ saptoAots snssig AVAOVLIA ac te at a Al. ai ae + + + O6 ore 0.16 @ ere, 6—0cel.ee,e¥pe.e es, 60 eels eioproned elIeosiq AVAOVNNVHY a= |) a Paououotione ar sca roo.o eieuodeg snpuides pile ||| | Baer acne Romo DT O eqejnyyeds “IPA E oo eSOoSIA vaeuopogy Je 4 + se eee Ce untasedees at as ab + + = + Ore) one 160 fe 1010 (610/00 67 0.50% uINPUIIOd umuriedsorpieg HVAOVAGNIdVS SFE | ee | tig fre Fo psi a oe | ee Beaty He | [29 || ropa abo undtc noc on topo-ot eyeaogo snuakey AVAOVULSVIaD) SIPS IEITIEIEIFIEP ISIE IEIZIEI Eee lele Blais lelaie (els lalalelelSIRleleasais ble B45 [=| aka Ht 5B |B & | & fay Bb |B /a2]s IF 218 & Coen | coon ES eel dems | eed | ese coe alel|ele eae Bl telet |e ge [8 |e =a = a | > 2 = | F panuuor SGNVISJ SODVdVIV5) AHL NOdN SHLAHdOLVWAAdS GNV SALAHdOGIWALd AO NOILNAINLSIGQ AHL ONIMOHS ATV] 2 CALIFORNIA ACADEMY OF SCIENCES (Proc. 4TH Ser. 188 eae TR a ae ee Siaioniee ba OI) state A) MUIR ONE, «|i Pshseesensitaerstsierenes acetates eijoyrurdivo “IVA VINO VPIS if | A i | et Pain| en Re Seite + afer AIT) aie al Oa | astaieuevene tedsuenes siesta tole ouaiare WNUPOTIOWIP WNIYSPA[PIY | ie al aT a ge a | a Sea : = BINA glist ction, crecieors ane ashen oe PP nee chee cen ee oe ane ee LGR! ORCL IMILID Wil eaitnee israats AR ks se ee ae fg fod) ce MH fo fag fl eee Seiya eee inten +| {+/—|+]+]+} J+] |4+]+ sta cde estat] Ollgseete Min ayercieaze asuapequeq wmid&sson ay po ee ety ee aie el eRe seer: Nanri ae Sere set alt tl teenetea ate peccrantea gts Bee erry Pisin ct et Oa MY ae ne TT a Fn pO NE eR ea see ot UM |g Py (A | A) a fade eect Ss ce 5 ot oh ae Bf ian gies As Ie a Pe ee eel Siteey ys ee 2 Bars (eh Vea an Vc eee rer CeNaeL neon AVAOVAIVIN eee re | eee ee eee Pie le le ale |e 3 5 ie} Q ea es % | 8 a #18 |% ® s 3 2 A B o a panurjuoy SGNWISJ SOOVdV1IV5) AHL NOdn SHLAHdMOLYWAAdS GNV SALAHdOGINALd AO NOILNATaYLSIG] AHL ONIMOHS AV STEWART—BOTANY OF THE GALAPAGOS ISLANDS 189 Vor. I] +++ Ten wate mena SOD UaEmn TaD Gutta psorqns eqoptavoury Sr OST once Be cree eprjao} vriopissed AVAOVAOTAISSVG Sra tented racet onc pyopmmpn vroumy, AVAOVAANANL Pte seaeereney teneteaceare arcana umyyoysayy wmotsed 4p AVAOVOIAAdAP PGE CON Le OTL VIPSWII,Ul BUIIOJ Sap He GST 21) Shae Muosslapuy PUIIO} 2 (coo ccc eyepnonar TUOYITE AVAOVITNDYNALS ie ee ce 2RaowATeW ssoes ce coemenc ne swny “IvA vITOJaVOIUOIOA ie ee vurdns (oc esourds Seta das onDpenopi no moray hs Propiquioyz ueurua A ysnor10qieny aqqearyeyopuy “PH Joupiey "yo dtoupies taddading ureyyey ajrzeuraq ry uopauiqy sosedeex) panwjuo) SGNVIS]T SODVdVIV‘) AHL NOdN SALAHdMOLVWUAdIS GNV SALAHdAOGINALT AO NOILNAINLSIC] AHL OSNIMOHS ATV [Proc. 4TH SER. CALIFORNIA ACADEMY OF SCIENCES 190 ++ ++ ++ ay[stonsh siete foreteteniveqemeniersueyaynyaneadeneotm tens 1TH a Ore erie oedeleeel elelene 09 Yeversexeretelnie.s vrosedyes eryundg $558) 6.60960) G/e).@ (0, aa eevegee '@.6 S.a5n (6 18 8 sndieoore]os ae CIEE Caan On Cn CC ei Ce snorjorsou WP cet can CLG ero sisuadedeyeS snoop AVAOVLOV) ashe diarageiareemeere “eyeqnorosey xIWJOIBTOS Cor eC Sat eC NT Yt Per . *eiodse BIlazyUs]NY AVAOVSVO'T Sept s anne vivseeseupkedeg Boneg AVAOVOIYVD ysnoroqien aqqearyeyepuy “PH seupiey “YD souprey) taddadjng uopauiqy sosederes) panuyuo) SGNVIS]T SODVdVIV‘5) AHL NOdN SALAHdMOLVWUAdIS GNV SALAHdOdINALG AO NOILNATMLSIG] AHL ONIMOHS ATV STEWART—BOTANY OF THE GALAPAGOS ISLANDS 191 Vor. I) eS 95 0 sel A legal el\ leek vos reese sdodoresdiscnf AVAOVYOVNOC) PAA aula dl: le \nceeisceeeteat aes giuiuascides SInOOTN AVAOVNOLSATASL 35 | ceeS |e ate aie |e Sta ofp llamas onset ences cat mccain esoulsoel BlIe]NouNnse] aft alk =} ae epeGeledohe!n.eheyee.eheeLesesbis)re)a,e,erelee, snyoala sndiesou0g AVAOVLAYAWNO) ee rs Refs lee (Ghee oneeoeeiacncns innameaepea mauris elem eli ag elie =| bomen ce aacorsaen : Sire treieaia AVAOVLYAS |p el alee le Palate GMa Wletencslen ssanecest neces oiueiy eioudoniay AVADVYOHdOZINY Bel eas fer fae es eee ence a a ETE ale ale alt ae Alp lececemer ee eco tare vcs ae en + © 0 Ghetele_ vu e,/0 vee esbhelbé lecete:-c- be exe) beets, einyed voydng AVAOVUHLAT slelelelelelelelelelzie/eleleyele ieee Bis (BIS lFlelSBlelelelslSiselel/Siele18 ja ls Bl |e|3 Be BR Pa Be ee ee es lieale is c ze ;: iS es Z panuyuoy SGNVIST SODVAVIVS) AHL NOdA SALAHAOLVWAAdS AGNV SALAHAOCINALY JO NOILNATYLSIC] AHL ONIMOHS AAV] [Proc. 4TH Ser. CALIFORNIA ACADEMY OF SCIENCES 192 stasof | same | ueurua | IaMo]L Inourfag | a1qesiyejopuy | ysnosioqien | a apauee (6 e16- 0)\4 ae ee Js)0) 616 lelene eens . ‘ds ee ee + wenn Ce er 2? unrpoytjaurdurrd Ns Fe Te i eet fan | eC recs Olct wniopiAied ‘lea wmurtaniod + +/)/4+)+ + + Seal Sen” i) cietentvsrezers JouIul “IvA uInyueTNosa umorsradooA’ : T ! I + + + + + + CO ee cece eee eee te ee were seen ceeeneedo ds + + + SH OA PN Si a ee tatis tareveveedeleneteeueraiehohenciiejetnie uinye nowuas uInIO“£ uot ahd + see Ce ‘ds Bie Ho] |) Yi evssevate atevaisiaicie:ese oie! sie esaisieieie ye et RINE vmyeq + — ORONO SCO ee ee LC) Cty tet Ke et umnuue umorsded aL He here +/+}—| | We ae Wlieracesarezescresste aseiezesetere chanrleleleseleyernyetersts HIsIoryy a siovede waeloteserstivenenekots taney Toakercrstavers TexooH] snqvoeg re) | 9 aed da rd om mo) CoN | [oy oval fale=I| lites Fal eden) ISIS IFT ISIZIFICIE ISIS ISISIF IEE lelele Ie (BITE le ]s ele le es (8/8 s/e)/a/eiF le BR ea |e cee:3 aia (BIS lelelFis (ele lale 5 eole A ie yee ce eal "T2118 is Es a eed |. Sa panuyuo 7 SANWIS] SODVdVIV) AHL NOdN SALAHAOLVWUAIS GNV SALAHAOGIVAL AO NOILAGIMISIC] AHL ONIMOHS aTaV] [Proc. 4TH Ser. CALIFORNIA ACADEMY OF SCIENCES 198 nn ere fetishes Oe tsrensiad et Peaccerener ya orohe icles gids vumooay, AVAOVINONDIG AE feslhats Hel |. [ereteereesereereeeeeeeeee sens s1pnp vuedoog Bviyenditls axavagn aPonetonegeu=Ceusiaetensratee vsoolynsy vIZaATeD apetetircseatenslel seneronearerweraay- ICR -puvraniod aT oN | (aN TE vsojid ‘1eA vloyiq vieided | | seeeeeee se Peis eres ede suas eiormuow vdoorg AVAOVIAVINHdONDS iia sige ay eveh BvaVsvele: onievendeageaal Hvetecsteas se es ereNePoue te “ds DE oyoi Fas she loss, eolipvore tole rw elsieleyshsysissmeretalis «da 7 ee ee ce +s) + wuntpoploseqioa ++ BRS On Gc Fitteeeeeee sess + gsua0}InG + aa + ao + + wil Nic gvananaravargherardowra. Wepeaudears cvgrerstecm ervey ear cic uinidiu ate eee eeeee Ob Sie: # eee 1c ole sie 616 eo 8 eee ole wie ee ‘ds ysnoloqien a1qesizejapuy saddading ureqzeq) apzewmaqry uopauiqy panuyuo7 SANVIS] SODVdVIV) AHL NOdN SALAHAOLVWUIMS GNV SALAHAOGINALG JO NOILNAINLSICG] AHL ONIMOHS ATV], STEWART—BOTANY OF THE GALAPAGOS ISLANDS 199 Vor. I] +++ ysnoroqreny a1qesiqejopuy *PH Jeupiexy "YO Joupiey daddading TEES) SGNWIS]T SODVdVIV!) AHL NOdN SHLAHADOLVWYAdS GNV SALAHdOdINALG AO NOILNAIMLSIG AHL ONIMOHS ATV ayyqerg sorpurg | eieDaletePewelcheacter nei sPetsieisd-den=veleteneteihat= ByROLIBAIP Sg |- =|), s|oo aeuecocoorocpmoror aan odGono essadsip eiedateiclRehshetekcneretoveneRotehsuelen est et eters tacalelicte Tineg RistenedenctapereNorategstereusr stay sVekeNe Rats rehstsre si[eseq Piialiog AVUOVIENY Rae Og ee eituind dIvA PSO}JUIUIO} i sofew oSequr[g AVAOVNIDVINVIG MEAD | oil Vere lNeseeahicce es etek keiseetesekteh seek atee uinpridsty wmisouresjoL Siiveiae Siva abensaiessioee Duepaherenane oe eqynowed erpony eI RSPR) ules cen cesetigarsicesi chert hse ies et pe lhe euvsedyyes vrorsn[ Siilayels ovatinalears daha s Bearer euetanied vraqdrpiq AVAOVHINVOW uo wulleg | ajrewaqy | uopsulqy | sosedryey) | paniwyuoy January 14, 1911. os il wn I & + 3S 3 a oy a CALIFORNIA ACADEMY OF SCIENCES 200 si michele erexelelsiercustetercickeccesLersteiatsys romue}4 aooovusadg tet We (hel UL ieminicceaceienceimieseeiecinlnn umidivoodéy umrunqpy ae flee sallicba lly | atieaies secarneqe an aaree eee Paes ees ene | Mee al eet ||| ete nwiielte'eiivvoreNete ns wwteue) ene, Caelegelieue eyeysnsue eiyoyoAsd : fa ae lak die alien otic 3 ctor atc thes g ADEN ErPOICl Be rule aa ON Ay Pelle slaogeeeias coeancater tac onteaar esrgerereano5 Z Bb a Ele n Setar atone Sct Meme anata es ae Baya wacooi 5 x RT eR ce A hh eas ah eG I. napacantoccsecntcaar cau Bineiedns 2 oe ane ty ee leg cee ee etenpanies Bell ST Sh | i el Fe tee ek ee) a We | NS I Hie rr ecicecece RCREReMCRERCRCRCRCRCY carey Cre CNA Becton eypisndied aa | Sl] ia tiara ao aye 2's tanehepe tere onscess eecsierekeisiece eyoyrared if Pert wry Me OPO TOINY GION ROR EMO SO CEO LIL Ceo eoytoed SHEA whe tala estes ete carezaiiey ators stsuouopsurqe PeUuIIOy SAF) Leth | dhe pl MUON 22) ences cc ASpee gen meat ee aueka oem lmnret aruiad ca NS Wt a OW ee ten aonseany] B/S le lg}ela|& PIRES Fle] 2] ¢ B\&|2 Be} alg a PR ese pe leaks ta Pa 5 el] ele} |) Wa eel ® . E panuijuo) SGNWIST SODVdVIV‘5) AHL NOdN SALAHdOLVWAAIS GNV SALAHdOCIYALG AO NOILAGIYLSIG AHL ONIMOHS ATAV STEWART—BOTANY OF THE GALAPAGOS ISLANDS 201 Vor. I] ueUUa A ysnoioqie Nn souref aiqeByejopuy poor | “PH Joupiey | "GQ rJeuprey | uevound | +++ + aeddading mreyeyo | sopey9 | ayeig aoqpurg | uozuLIeEg + +++4+ ayzemeqyy | uopsuiqy | sosedeyery | CLES OE LRT BCS OH snyeue, snddedojdy ci uns ie ae eI[Ojavisiwiayie vIsoIqUIYy Mi hAUMIAOO Homa aolao sepiozhuoo wnqPIe8y soz domo agcdddo eR oeeonooon umdevooso1w PAE oN aol ele seplodiesoone, wnuiadsoyyuPoy AVLISOdNOD Fp saonpagoowIpubnoa casas Wonun]g PB[OAsROg AVAOVINAGOOL) Harm pancantosonto dasa eURIOyO eieqoy AHVAOVIONVdWYV’) Br icceyeteraqevsysterentts layenetetore reser arms snsoj]iA soforg souoedhoupUduooDSdaDot eyyuPILYyD voIpsowoyy DHHOUDD UU DOO Doon op O60 0 wmnyepi0o uMTOWT RAMP AMOTO ROMOO Heb Moron ono oo odeg vytqmong FeO MOO OGD aaa NGG cD sue3na snqnsy9 AVAOVLIa ANON?) panuyuoy SGNVISJ SODVdVIV‘) AHL NOdN SALAHdOLVWUAdS AGNV SALAHAOGINALT AO NOLLNAIMLSIC: AHL DNIMOHS ATV L CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. 202 +. ps =e 4) SiL8 Jeri6,/6,6,01 0/.6,0',0 10),0:30! 16) /61,9.(0:101,0)'8 foveile's 6, epidsiy vI[soUgy ae + CaCO ORC Ca CORCORAN CnC Ceci suodas ee | ANS LI tsiietebentesie seve covielerejfeie| lets te leicejie jee isis Jai e): sis suesdoajour VILALTA afe7.e| 6 10's 076! @ ‘68 6 )0\u6)6 86) 0088 60.8 ems le)e Pyooalo eydipoq —_ uinq{jisnd +++ OTe Ae Foss easels erst ecuep sie agys winjoala wnyjeyyursAayo, whe [040 16ife .0, (ol Tok (ofibyie;ec:iarieie(6velepe;teiee esnyip vyjexoug ae) e apa(eis ot (ele ee eee te 8 eee ee evar nie st[nevomuay + + of +++ +++ | +++4+ 14+ 3 e Se Mate PT a Pa)" capita peacatnie oereaiatsiecetiaen sit ace apeare 8 11212018 : Galt Mh *) oigiertee thie tine VULISSIYSNSUB “IVA voOvIBUIg ee Ne TL VA LA ci] siivselistote Mette xe: eupits is sakes toils, eXer"evers".s) 0} explo siiepnyid sliByoorg ueuIUus | JaMoL saddading, weyieyy uo W;WuLIeEg azewmaqry uopsulqy sosedeyery) ysnosoqieny “GD Joupres) a[qesiqejapuy panuyuoy SANVIS]T SODVdVIV) AHL NOdA SALAHAOLVWAAdS UNV SALAHAOdINALY AO NOILAMIULSIG] AHL ONIMOHS AVL, STEWART—BOTANY OF THE GALAPAGOS ISLANDS 203 Vor. I] a tea lite Steal lista stad [ait [tae lp Staci in eyopolsey eyoeyoodry Se) es oa A fl ne ictal | | ROSES TCS eC Gr SO dae snpyyeuurd sndivo0eT Ne eS ee eae alt SN CIE Siicccoomconoronoccronenougdaany suadar0id ts pr te ee fee oe eso anon onatucc eqn Sale Te OSS lal) S| Hovoepe copansoanoumoconepoDe stfovs3 vioSovf =i hip BOGR igs Coto ue oskb.c. O04 Capa eptjenbs vruoziweyy mf foes pees | Makes aRractenese de ccehentceS ssc oishs wumnqy]e-09}N] wunieydeuy =| ere NS al J bnetomonsnoccanamcctudcoacc. stjuapiq emaaryy ef | | tree eee ee ret ee sien er tear is PT a VP ANE NWWsoboscooqnaocosononocoD cc o.nccseSo arts a of SCO aneory umioyedng eT ac Va er) Goa nba Ue be MR Ve (a Uo ue gM es snsojusuloy “IeA |e ae a ee ee oc snijoymuoy alien Be =) srerereseinisas & reperdovenersin else era cr snyoytury oe snfnosniqr]s ‘rea Te oe ee ee eo snyoyiour] uorsg ey ese tle lela (sls lela lee lala ele z ¥ & 3 z1o g 1/8 8 aoe les o Saale panuyuoy SANVIST SODVdV1IV‘) AHL NOdN SALAHdMOLVWAAdIS ANV SALAHdOCINALd AO NOILNATALSIC] AHL DNIMOHS ATV a\'s} Sie: 6 /efayelcheliey oiielleiiers [ole lcueteteleieleletaisveiehets’s BSIAID af my ee) (eee ecco ot nO ck norOr hon Caco nn Onc suaiinoap (Proc. 4ruH Ser. ak an © bla aie ele Be ese Bie ee we Bie Whletsire. sieiip 4 lelitfe eve: eyeps0o + @ 0 c40 ee) ere 6 0) 4,0, 4,0 0 0se 01 0.8 Be eyeiqrys “ICA ete 2 ore be cehe wa 6e O04 bse 8 6 ee 0.8 8 eRe 0,108.0) a0 ere nineg de Cede Onl eee -cele elec ce hee oe soproyAqoriye S66 oe Oe ele Oe 0 88 Ore wie 08 8.8 Sie eee a8 viodse CRO Ce ON ec Cn auc re arc eva sluyge RISaTeOG aa ait ab eee Step | ta | me | fchecstenciene cackencr sa cre hegets "+ a[elepni wn Aydor0g ae aD Pe | ec on Roa Cg nod Sars aN elpoymmusy +++4++ CCA ai ee Faso il becomodon cawudnosoomddenonbas vso1enbsqns SOOO OD ct amcun np Ot omarto nictte t8 fn IOUT] oo | ee ere ae ve SURI VEGI Al | fsake earsenete cc tese heheh caer acacia cas Iayooyy =e) ce FUN) |e fe reed ea GV We carers aca ie ca Se Pace muosssiopuy siqoaq | aL CALIFORNIA ACADEMY OF SCIENCES fe + uPUIUa MA | amo], | snourf£ag | ysnoioqieny | stasof | souref wmeyyeyD ajszeueqiry uop3uiqy ‘PH Joupseg a] qesryejapuy panwuyuo7) SANVIS] SODVdVIVD AHL NOdN SAILAHAOLVWAIdMS UNV SALAHAOGIMAL AO NOILNAINLSIG AHL ONIMOHS ATV 204 205 STEWART—BOTANY OF THE GALAPAGOS ISLANDS Vot. I] *puvysy uojdweyD 1 ne ee eee ee en ueurus ysnoroqie Ny ges yLJepuy PH Joupieg qo Joupiey saddading, ees Jima fours ee, See ee ee Perey onan BaD eee specter eee eiouoy ““5-17|Indefatigable |'Common and|Rare Common, mid- (South) fresh, at 600 ft, sea on Oe ‘* 18-20 Indefat. (NE.) S ‘* 21-30)\Indefat. and S. Seymour Kansas University, November 29, 1910. September 28, 1911 322 CALIFORNIA ACADEMY OF SCIENCES [Proc. 41TH Ser. EXPLANATION OF PLATE XX Agraulis vanille Linn. é Old Colony Settlement, Albemarle Island, April 24, 1906. Agraulis vanille Linn. 6 Under surface. Tagus Cove, Albe- marle Island, March, 1906. Cupido parrhasioides Wallengren. Chatham Island, October, 1905. Cupido parrhasioides Wallengren. @ Chatham Island, Oct. 1905. Cupido parrhasioides Wallengren. g Under surface. Chatham Island, October, 1905. Eudamus galapagensis Williams. 6 Wreck Bay, Chatham Island, October, 1905. Phlegathontius leucoptera Roth. and Jordan. 9 Wreck Bay, Chat- ham Island. Raised from caterpillar caught March 12, 1906. Phlegathontius (Protoparce) Calapagensis Holland. @ Light phase. Charles Island, February, 1906. Phlegathontius (Protoparce) Calapagensis Holland. ¢@ Dark phase. Tagus Cove, Albemarle Island, April 12, 1906. Phlegathontius (Protoparce) Calapagensis aberration nigrita Roth. and Jordan. 6 Tagus Cove, Albemarle Island, April 12, 1906. J Dilophonota obscuris Conformis Roth. and Jordan. 6 Tagus Cove, Albemarle Island, March, 1906. Proc Cat Acan. Sci 4.7 Ser VoL Cama, seed: and a species of yi Urticaceae Acacia, Croton , Corda, Cactaceae ,Bursera, Gossypium miRNA KER l THE WEATHER (SH) SIDE OF INDEFATIGABLE ISLAM 1 Dry Zone-to 200L., 2Llight Green Zone - to Z00Lt., F [WILLIAMS ] PLATE XXI : Light color. Teller ferns, 6rass,and Lichens; or perhaps wees covered with lichens ~ Browmsh, no trees (2) zones not s0 well detined here _--» Gpen area = SS — PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FourtH SERIES Vor. I, pp. 323-374 January 17, 1912 EXPEDITION OF THE CALIFORNIA ACADEMY OF SCIENCES TO THE GALAPAGOS ISLANDS, 1905-1906 IV THE SNAKES OF THE GALAPAGOS ISLANDS BY JOHN VAN DENBURGH Curator of the Department of Herpetology CONTENTS Pirates XXII-XXX PAGE INTRODUCTION . ; : ‘ P 0 5 5 ; : F ‘ 324 PrEVIOUS COLLECTIONS AND STUDIES 3 é : ; A ‘ 324 Tue Genus Dromicus Breron . ; F : : F ; : 327 Key To THE GALAPAGOS SPECIES. : z d ' ; ‘ 329 Tue MATERIAL FOR THIS STUDY . : : : F i 329 ORIGIN OF THE GALAPAGOS SNAKES . : 4 ‘ ; b ; 331 SuccEsTIoNs To Furure STUDENTS . F : : ‘ : : 336 DISCUSSION OF THE SPECIES . : F Z 5 : ; : ; 336 Dromicus biserialis (GUNTHER) . ; ; F : ; ’ 336 Dromicus hoodensis, new species . . : ; : : : 338 Dromicus dorsalis (STEINDACHNER) 5 3 : : ; 5 341 Dromicus occidentalis, new species . 3 i : : ‘ : 347 Dromicus occidentalis hellerit, new subspecies F F : ‘ 349 Dromicus slevini, new species . F : é R ‘ ; i 351 Dromicus steindachneri, new species . : ‘ . 5 5 353 Hydrus platurus (LINNAEUS) . + : S . ; : d 355 EXPLANATION OF PLATES é : : : : " Y ‘ ; 356 January 15, 1912 324 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H SER. INTRODUCTION In reporting upon the snakes secured by the Academy through its expedition of 1905 and 1906, I wish first of all to express my appreciation of the energy and care of my assistant, Mr. Joseph R. Slevin, upon whom, as chief herpet- ologist of the expedition, rested the responsibility of gathering and preserving the collection which has made this paper pos- sible. I am indebted to him also for the counting of many scales. To Mr. E. S. King, and to other members of the expedition who aided in the collection of reptiles, my thanks are due. Professor Charles H. Gilbert, as so often in the past, has aided me by kindly permitting me to make use of speci- mens in the collection of Stanford University. From Dr. George A. Boulenger I have received, regarding certain speci- mens in the British Museum, information which has been most useful. All measurements are given in millimeters. The numbers by which specimens are designated are the serial numbers of the reptile collection of the Academy, except such as are pre- ceded by the letter S. These latter are the numbers attached to specimens in the collection of Stanford University, and refer to the register of its reptile collection. The sea snake Hydrus platurus is here first recorded from the Galapagos. The following snakes are described as new: Dromicus hoodensis Dromicus slevint Dromicus steindachneri Dromicus occidentalis Dromicus occidentalis helleri PREVIOUS COLLECTIONS AND STUDIES It is probable that the presence of snakes in the Galapagos Archipelago was first recorded by Dampier, who, in his Voy- ages, mentions green serpents seen there in 1684. Delano, Porter, and Darwin refer to them briefly in their Narrative and Journals. Darwin, I believe, was the first to carry back to Europe a specimen of this snake. It was caught on Charles Island, and Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 325 Bibron considered it identical with a Chilian species. It was so regarded until 1860, when Ginther, in the Proceedings of the Zoological Society of London, pointed out certain differ- ences between the mainland and the Galapagos snakes, and named this Charles Island specimen Herpetodryas biserialis. In 1869, Peters recorded a specimen in the Museum of Stockholm, collected in the Galapagos by Dr. Kinberg, as identical with the mainland Dromicus chamissonis. Gtinther in the Zoological Record for 1869, remarks that he “can con- firm Professor Peters’s observations, having now seen a series of examples of this snake from these islands. There were two varieties, one very similar to the common continental form, the other identical with the snake described by him from a young specimen under the name of Herpetodryas biserialis. Some examples were intermediate between the varieties, so that there is no doubt about their specific identity. The syn- cranterian character of the dentition is not well developed in this species.” The Hassler expedition secured no snakes in the Galapagos Islands, but one was seen upon Jervis Island, in June, 1872. Stiil later, Dr. Steindachner secured for the Vienna Museum five snakes which Dr. Habel had collected in the Galapagos Archipelago in 1868, and which, he says, are the specimens to which Dr. Giinther referred in his note in the Zoological Rec- ord for 1869. These specimens showed two types of colora- tion—spotted and striped—and Dy. Steindachner regarded them as two varieties of the continental Dromicus chamissonis. The spotted form he called Dromicus chamissonis var. dorsalis, while the striped specimens were named Dromicus chamissonis var. habelii. These snakes were said to have been found on Indefatigable, Hood, Charles and Jervis islands; but the gas- trostege counts given by Dr. Steindachner, and his description of the post oculars and temporals, differ from the conditions found in the snakes of Charles and Hood islands to an extent which enables us to say that his specimens must have come from Indefatigable or Jervis. No other names have been proposed for Galapagos snakes. As the years have passed, and snakes have been found on Charles, Hood, James, Jervis, Barrington, Indefatigable, Albe- marle, and Narborough islands, authors have sometimes * 326 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4rH Ser. regarded them as identical with the mainland Dromicus cham- issonis, sometimes as one or two distinct varieties (spotted and striped) of this mainland species, sometimes as a distinct species, D. biserialis, with or without a subspecies, D. biserialis habelii. As Garman put it, “there is nothing in the published evidence to show that the striped form, the spotted form, that with two postorbitals, and that with three do not occur amongst the individuals of any of the localities inhabited by this snake. Giinther’s type has three postorbitals and is spot- ted, Dr. Baur’s specimen has three postorbitals and is striped, and Steindachner’s varieties, both striped and spotted, have but two postorbitals.” Even as regards the generic term to be applied to these snakes, there has not been agreement among herpetologists. Giinther at first placed them in the genus Herpetodryas, but later followed Peters in referring them to the genus Dronticus of Bibron. Here they have been placed also by Steindachner and Boulenger. Cope, in 1889, applied to them the generic name Opheomorphus Fitzinger, but Garman has shown that this is a synonym of Liophis Wagler, being founded on the same type. Garman reverted to Fitzinger’s Orophis of 1843— the type of which he states is Coronella chamissonis Wiegm.— because he held that the species of the Galapagos Archipelago of Chile, and of Peru differed generically from the West Indian species, which he retained in Bibron’s genus Dromicus. Still later, Cope divided all these snakes into three genera: Dromicus Bibron, with no scale-pits; Monobothris Cope, with one scale-pit; and Alsophis Fitzinger, with two scale-pits. Monobothris Cope has as type Dromicus chamissonis, and is therefore a synonym of Fitzinger’s Orophis which was based upon the same species. Stejneger has called attention to the fact that Bibron’s Dromicus, 1842, is preoccupied by Dromica Dejean, 1826, and has revived Fitzinger’s Leimadophis for the species which normally have no scale-pits; but the recent ruling of the Committee on Nomenclature of the International Congress sanctions the use of the name Dromicus. Leima- dophis therefore must revert to the synonymy. We thus have left three generic names—Dromicus Bibron, 1842, based upon a West Indian species without scale-pits; Orophis Fitzinger, 1843, established upon the Chilian species Vor. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 327 with one scale-pit; and Alsophis Fitzinger, 1843, the type of which is a West Indian snake with two scale-pits. The only character which has been held to distinguish these genera is the number of scale-pits. However, since the snakes of the Galapagos Archipelago are certainly congeneric, and since it will be shown that they have scales with two or one or no pits, there seems to be no good reason for recognizing more than one genus for all these snakes—West Indian, Chilian and Galapagos—which agree in every other respect. Any other course would mean the establishment of genera which were in no sense natural groups; for the Hood Island snakes are certainly more closely related to the other Galapagos serpents than they are to the West Indian species which have no scale- pits. It would seem that as differentiation has proceeded, certain of the species in the Galapagos have lost their scale- pits, as others have in the West Indies. Tue GENUS Dromicus BIBRON 1842, Dromicus (not Dromica Dejean, 1826) Brsron, in Sagra’s Hist. Fis. Pol. Nat. Cuba, IV, Rept., 1842, p. 133 (type Coluber cursor) ; Bou- LENGER, Cat. Snakes Brit. Mus., II, 1894, p. 118. 1843, Alsophis, Firzincer, Syst. Rept., 1843, p. 26 (type Psammophis areiilonsis Schlegel) ; STEJNEGER, Report U. S. Nat. Mus. for 1902, 1904, p. , 1843. Leimadophis, Firzincer, Syst. Rept., 1843, p. 26 (type Coronella a aa regin@) ; STEJNEGER, Report U. S. Nat. Mus. for 1902, 1904, p. ; 1843, Orophis, Frrzincer, Syst. Rept., 1843, p. 26 (type Coronella cha- missonis Wiegm.). GARMAN, Bull. Essex Inst., XXIV, 1892, p. 86. 1843, Calophis, Frrzincer, Syst. Rept., 1843, p. 26 (type Herpetodryas cursor). 1854, Teniophis, Grrarp, Proc. Acad. Nat. Sci. Phila., 1854, p. 226 (type T. tantillus=D. chamissonis). 1862, Haliophis Core, Proc. Acad. Sci. Phila., 1862, p. 77 (emend.). 1882, Alophis, Staut, Fauna Puerto-Rico, 1882, p. 70 (err.). . 1884, Ocyophis, Corr, Proc. Amer. Philos. Soc., XXIII, 1884, p. 491 (type . ater). 1887, Halsophis, Corr, Proc. U. S. Nat. Mus., X, 1887, p. 439 (emend.) ; Corr, Trans. Am. Philos. Soc., XVIII, 1895, p. 201. 1894, Liophis (not of Wagler, 1830), BouLENcER, Cat. Snakes Brit. Mus., TI, 1894, p. 126 (part). 1894, Monobothris, Core, Amer. Nat., 1894, p. 841 (type Dromicus chamissonis) ; Corr, Trans. Am, Philos. Soc., XVIII, 1895, p. 201. All of the land snakes of the Galapagos Archipelago agree in their dental and hemipenial characters. The maxillary teeth vary from ten to twelve in number, followed, after an 328 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. interspace, by two larger ones. Thus, counting all sockets as well as teeth :— No. 11935, Dromicus hoodensis, from Hood Island, has 12 and 2. No. 11800, Dromicus hoodensis, from Hood Island, has 12 and 2. No. 11926, Dromicus hoodensis, from Hood Island, has 12 and 2. No. 11930, Dromicus hoodensis from Hood Island, has 10 and 2. No. 10782, Dromicus dorsalis, from James Island, has 11 and 2. No. 10483, Dromicus dorsalis, from South Seymour Island, has 10 and 2. No. 11488, Dromicus occidentalis, from Narborough Island, has 11 and 2. No. 10281, Dromicus occidentalis helleri, from Brattle Island, has 10 and 2. No. 10617, Dromicus steindachneri, from Jervis Island, has 11 and 2. The hemipenes of Dromicus hoodensis (No. 9336) from Hood Island, of Dromicus slevini (No. 12216) from Duncan Island, and of Dromicus dorsalis (No. 10483) from South Seymour Island, all are divided, with furcate sulcus, calyculate, spinous proximally, and with no apical disc. They agree in every respect with the figures given by Cope of these organs taken from “Monobothris’ chamissonis, “Alsophis’ angulifer and Dromicus parvifrons of Peru, Cuba and Hayti. Scale-pits do not occur in all the scales of any specimen from the Galapagos. When they are present, they are most constant in the scales in or near the region of the lateral stripe and on the upper surface of the tail. Most careful examina- tion has failed to disclose any trace of pits in any scale of any of the Galapagos snakes having fewer than one hundred and ninety gastrosteges. The Hood Island and the Charles Island species also normally have no scale-pits; but long search on the thirty-six specimens at hand from Hood resulted in the discovery of a single scale with one pit. Excepting the species from these two islands, all of the snakes of the Galapagos with Vout. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 329 more than two hundred gastrosteges bear scales with two pits. They also have scales with no pits, and usually others with one pit. In some cases large scales on the tail have three or four pits. While these pits, therefore, are not of generic value, they are of great use in the separation of species, as shown in the following: KEY TO GALAPAGOS SPECIES OF DrRomIcus a.—No scale-pits. b.—Gastrosteges more than 195 (203-214). c.—General coloration in spots; scales in 19 rows. Charles and Gardner-near-Charles. Dromicus biserialis.—p. 336. c.2—Striped, the stripes fading out posteriorly; scales in 17 or 19 rows. Hood and Gardner-near-Hood. Dromicus hoodensis.—p. 338. b.2—Gastrosteges fewer than 195 (169-183). cc.—Postoculars two; no longitudinal light stripes. Duncan, Albemarle, Narborough. Dromicus slevini.—p. 351. cc.2—Postoculars normally three (rarely two) ; longitudinal light stripes. present. Jervis, South Seymour, Indefatigable. Dromicus steindachneri.—p. 353. a.2—Scale-pits present. bb.—Gastrosteges more than 210 (213-252), ecc.—Gastrosteges usually not more than 232 (213-236). James Jervis, Barrington, Indefatigable, South Seymour. : Dromicus dorsalis.—p. 341. ccc.2—Gastrosteges not fewer than 236 (236 to 252); prominent light markings on nape spots or transverse blotches. d.—Usually striped; light nuchal blotches and a series of dark spots on tips of gastrosteges and on Jower lateral scales very distinct. Narborough. Dromicus occidentalis.—p. 347. d.2—Spotted, without longitudinal light stripes; no series of definite rounded blackish spots on lateral scales of first and second rows; light nuchal markings less prominent. Albemarle and Brattle. Dromicus occidentalis helleri—p. 349. bb.2—Gastrosteges fewer than 210 (178-201). Chile and Peru. Dromicus chamissonis. ’ Tue MATERIAL FOR THIS STUDY It will be seen that I have recognized seven kinds of land snakes from the Galapagos Archipelago. This has been made 1 There can be little doubt that more than one species occurs in Chile and Peru. The wide range in the number of gastrosteges would indicate this, and Dr. Boulenger, who most kindly has examined the scale-pits in the specimens in the British Museum in response to my request, writes me that most of the Chilian and Peruvian specimens have scales with single pits, while those from Chiloe have scales with two pits. These specimens from Chiloe doubtless represent a distinct species, as yet unnamed. 330 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H SER. possible only by the large number of specimens secured. The collection included ninety-eight snakes from these islands, and I have also had the privilege of examining eight in the Stan- ford University collection, making, in all, one hundred and six specimens, distributed as follows: 13 Keo(< Meu pei aT ners DOE eEn cre PcG Ofc 36 Indetaticabletar ss rcrtertetoriterretaeisterrens 24 Barrington. 2a cere eis ecient oe ee 15 Uiekecvec A ance Amin ca Modo cod mioms.ddo.cinio INarborotigh: <5 jajs,steub store eneyepete level ersrenelle rs South Seymour jasc roe eros Jervis) 243 vepeciesectdtre teaitebotaus Biker ope cNerreIeys Brattlerz: tia tcasisie siveicvae coterie aeieie Gardner-near-Elood =). sec cere coe Gardner-near-Charlesi oi 1s. )21<1- t-te ai Duncans eos cicierre cy ncrmeed a aera Gowley, Mt. “Aillbemarle 3-5-4 402266 Cape" Berkeley, “Albemarle: ope eee eee eet DD BUNT 00 Although this material seems large, it is quite inadequate for the final settlement of many of the questions which present themselves. The series from Hood is the only one that really is satisfactory. The Indefatigable series might at first seem so, but one of the species found on that island is represented only by a single specimen; and the twenty-three examples of the other species are not enough to furnish a satisfactory explanation of the presence of both spotted and striped styles of coloration. The numbers secured on the other islands are, of course, still less satisfactory, especially when one recalls that we have two distinct species from several of the islands. It is probable, too, that larger series from many of the islands would enable us to recognize specific or subspecific differences which are now hidden by individual variation. Thus, the snakes which I am forced to group together as Dromicus slevini may very well represent at least two different races. Similarly, the snakes of James and Jervis may be found to differ from those of Barrington and Indefatigable, as is pointed out under the head of Dromicus dorsalis, and those of Brattle possibly will be found to be not identical with those of northern Albemarle. The solution of these problems, how- Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 331 ever, must await the gathering of larger series from all the islands except Hood, and perhaps Charles. No snakes ever have been taken on Culpepper, Wenman, Abingdon, Bindloe, Tower, or Chatham islands. One of the residents of Chatham told Mr. Slevin that snakes were not uncommon there, but careful search failed to bring one to light. They must now be quite rare on Charles; for no mem- ber of our expedition saw one on Charles Island itself, although one was secured on the close-lying islet known as Gardner- near-Charles. rs ORIGIN OF THE GALAPAGOS SNAKES The closest relatives of the serpents of the Galapagos Archi- pelago are a number of distinct species native to the Bahamas, Greater and Lesser Antilles, Costa Rica, and all of South America—species which Boulenger includes in the genera Dromicus and Liophis. Whether or not all of these species actually belong in. the genus Dromicus cannot be positively stated until the hemipenial structure of each has been exam- ined. The results of such an examination, however, cannot be expected to affect the truth of the statement that the Gala- pagos snakes have very close relatives throughout the West Indies and South America. This being true, the snakes of these localities must have had a common origin. Either the West Indian and Galapagos snakes have been derived from South America, or else all must be descendants of species which, in a former geological period, occupied a great central land-mass which has sunk below the level of the sea, leaving mere remnants in Central America, northern South America, the Antilles, and the Gala- - pagos. Much may be said in favor of each of these theories. I believe that the data are not yet at hand which will enable us to choose between them. Either view implies a former land connection and a conti- nental origin of the Galapagos ophidian fauna. I cannot ‘bring myself to share the opinion of those who believe that the fauna of the Galapagos has reached these islands by the more or less accidental agency of the winds and ocean currents. The various species must have spread slowly over some conti- ° 332 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4rm Ser. nental mass with which the Galapagos were connected or of which they formed a part. When the Galapagos finally became separated from the rest of the world, it is probable that most or all of the present islands remained for a time united. The northern islands must have been the first to establish an independent existence, and it is possible that their separation may have occurred before snakes reached the Galapagos, and, therefore, before the old continental bridge was broken; but I think it more probable that snakes once inhabited these islands also. Cul- pepper and Wenman islands are, of course, unfavorable for the continued existence of snakes. Just why they never have been found on Abingdon and Bindloe is indeed hard to under- stand. While all of the snakes of the Galapagos Archipelago are closely related, they nevertheless are of two distinct types. These are the small snakes with no scale-pits and fewer than one hundred and ninety gastrosteges, and the group of species with more than two hundred gastrosteges. These two groups I believe to be the descendants of two species which originally occupied the Galapagos. My chief reasons for this opinion are the absolute distinctness of the two groups, and the fact that representatives of both have been found upon the same islands. The snakes with more than two hundred gastrosteges fall naturally into three subgroups. These are: first, the snakes of Charles and Hood; second, those of Narborough, Albemarle and Brattle; third, those of James, Jervis, Indefatigable and Barrington. The first of these subgroups is the most distinct. Differenti- ation has progressed much farther on Charles and Hood islands than elsewhere in the archipelago. Therefore, we may believe that these southern islands were separated from the central ones before the latter were divided one from another. Tlie snakes from Charles and Hood islands are very closely allied. They agree in all essential characters except color. They alone of the larger Galapagos snakes lack the scale-pits, and both have the same number of gastrosteges. Differenti- ation could hardly have occurred along lines so absolutely parallel in two unconnected islands. We are therefore led Vou. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 333 to believe that Charles and Hood islands were connected, and formed a single island for a long time after their separation from the more northern or central islands. The snakes of the two Gardner islands agree in every detail with those of the larger islands to which these are adjacent, so that the separation of the one Gardner from Charles, and of the other Gardner from Hood, must have occurred still more recently. The second and third subgroups are much more closely related to each other than to the first. This may be considered to indicate that all of the central islands from Narborough to Barrington and from James to Brattle—with the possible exception of Duncan—remained connected for a considerable period after the separation of the northern and the southern islands. The distribution of the second and third subgroups, and of D. slevini and D. steindachneri, indicates that there occurred at a still later date the separation of this central land into two large islands; an eastern, including the present James, Jervis, Indefatigable and Barrington Islands; and a western, of which Narborough, Albemarle and Brattle formed parts. The more recent changes are much less clearly indicated by the ophidian fauna, but certain color-differences render it probable that Narborough became separated from Albemarle before breaks in the eastern island occurred, first between Barrington and Indefatigable, then between James and Inde- fatigable, and lastly between James and Jervis. The snakes of Albemarle are at present known only from two specimens—one Dromicus slevini from Cowley Moun- tain, and one Dromucus occidentalis hellert from Cape Berkeley. Under such conditions little can be deduced as to the past history of this island without the use of evidence furnished by other groups of its inhabitants. This evidence I do not now wish to use; for I believe more accurate results can be attained by attempting to read the story of each group sep- arately, and then comparing results. The mixing of evidence here, it seems to me, would be only less confusing than the jumbling together of data derived from distribution, geology, paleontology, and ocean-soundings. Each should be worked 334 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4H Ser. out separately before comparing results, in order that one may serve to confirm or disprove the other. If we have read this story of the snakes correctly, there is nothing in the least suggestive of an unconnected group of volcanic islands thrust independently above the surface of the ocean, to become the home of such animals as might reach them through more or less accidental or occasional agencies of dispersal. Instead of telling of the elevation of new islands, the evidence points to the gradual depression and partial sub- mersion of a more extensive land-mass which must have had direct or indirect connection with continental America. When we consider the snakes from the various islands as regards the style of their coloration—whether spotted or striped—we find an interesting fact. On almost every island only one style of coloration is present. Thus, all the snakes of Hood, James, and Jervis are striped; while on Charles, Albemarle, and Brattle only spotted snakes have been found. But when we come to Narborough, Indefatigable and Bar- rington islands, we find that each island has both spotted and striped snakes. Why should a difference of coloration so constant on other islands be inconstant here? We have seen that the snakes of Charles and of Hood are alike, except that those of Charles are spotted while those of Hood are striped. If these two islands should now become connected for a time, we might expect spotted snakes to wander to Hood, and striped ones to appear on Charles. If these islands again became separated, we should find both spotted and striped snakes on each island; but if the connection had been short, we might expect a majority of the snakes of Charles, and a minority of those of Hood, to show the spotted coloration. Fifty-three per cent of the fifteen snakes from Barrington are spotted. Seventy-four per cent of the twenty-three speci- mens from Indefatigable are striped. More numerous speci- mens might change the proportion and show that the sug- gested explanation is quite wrong, or that differentiation is now for the first time developing between the Indefatigable and the Barrington snakes. The parallelism between the con- ditions actually found on Barrington and Indefatigable, and the conditions which we might expect to find upon Charles Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 335 and Hood, should they now become connected and later sep- arated again, is strongly suggestive. It might be thought to point to an elevation and depression of Barrington and Inde- fatigable subsequent to the general depression of the archi- pelago. This view might be strengthened by the fact that all of the snakes of South Seymour Island are striped. Certain slight peculiarities of coloration, however, distinguish most of the Barrington Island specimens. from those of Indefat- igable. With respect to these peculiarities, the striped snakes of Barrington differ from the striped snakes of Indefatigable, and agree with the spotted snakes from their own island. Similarly, the spotted snakes of Indefatigable differ from the spotted snakes of Barrington, but agree with striped specimens from Indefatigable. Therefore, we must regard this as a case of dichromatism, occurring in the snakes of these two islands; but if similar proportions hold in larger series, it will be evident that specific differentiation has already begun, and may ultimately lead to the formation of spotted and striped races here as it has on Charles and Hood and on Albemarle and Narborough islands. The following diagram will serve to show the probable relationship of the snakes of the Galapagos. -D. cceidentaulrs Do. hetlerz Tames yJerv e's N_Barring ton D. bisereadrs D.Roodengis Dsieandachnert D.slevini 336 CALIFORNIA ACADEMY OF SCIENCES _ [Proc. 47H Serr. SUGGESTIONS TO FUTURE STUDENTS Future collectors in these islands should strive to secure specimens of the snake of Chatham Island, if such there be. Doubtless, it will prove to be a most interesting new species. Duncan Island is one of the most difficult to understand of all the islands of the archipelago. Its snakes are represented in collections only by a single specimen of D. slevini, although there can be little doubt that a larger species, probably with two scale-pits, remains to be found there. Other specimens of D. slevini have been seen on Duncan Island; and, since these agreed perfectly in coloration with the type, it is almost certain that additional specimens from Duncan, Albemarle, and Nar- borough will show that more than one species is here referred to D. slevini. Many more specimens of D. stemdachneri also are needed. Much remains to be learned of the larger snakes of Albemarle, which now are known from only one or two specimens. Dr. Boulenger writes me that the British Museum has a specimen with 222 gastrosteges, which is said to have been collected at Tagus Cove. I am inclined to doubt the correctness of this label; but if no error has crept in, there must be more than one species with two scale-pits in this island. The question then arises: Is there in Albemarle a distinct race of snake on each of the five principal mountains, as there is of tortoise? The answer must be based on many specimens yet to be collected. The question of the necessity of further division of Dromicus dorsalis also remains for future collectors to solve. DISCUSSION OF THE SPECIES Dromicus biserialis (Giinther) CHARLES ISLAND SNAKE 1860, Herpetodryus biserialis GUNTHER, Proc. Zool. Soc. London, 1860, p. 97 (type locality Charles Island). 1869, Dromicus chamissonis GUNTHER, Zool. Record, 1869, p. 115 (part) ; Bou.encer, Cat. Snakes Brit. Mus., II, 1894, p. 119 (part). 1876, Herpetodryas dorsalis, STEINDACHNER, Festschr. Zool.-bot. Ges. Wien., 1876, p. 304 (err). Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 337 1892, Orophis biserialis GARMAN, Bull. Essex Inst., XXIV, 1892, p. 85 (part). 1903, Dromicus biserialis biserialis HELLER, Proc. Washington Acad. Sci., V, 1903, p..93 (part). Diagnosis.—No scale-pits; scales in 19 rows; gastrosteges 209; urosteges 108 to 110, all paired; postoculars three; tem- porals usually 2+-2; spotted. Type.—British Museum. Charles Island, Galapagos Archi- pelago. Charles Darwin. 1835. Distribution —Charles and Gardner-near-Charles islands, Galapagos Archipelago. Material.—Only two specimens of this species are in collec- tions. These are: the type, a young specimen from Charles Island, preserved in the British Museum, and one female speci- men from Gardner Island—No. 9448 of the Academy collec- tion. Description of No. 9448—Head rather long, with flattened top and rounded snout. Rostral plate large, a little broader then high, hollowed below, and bounded behind by internasal, anterior nasal and first labial plates. Plates on top of head are: a pair of internasals, a pair of pre- frontals, supraocular and part of preocular of each side, a frontal, and a pair of large parietals. Internasals much smaller than prefrontals. Frontal longer than parietal suture. Anterior and posterior nasals distinct. Loreal well developed, longer than high. One large preocular with a very small one below it on each side of head. Postoculars three. Temporals two followed by two or three. Eight superior and ten inferior labials, sixth upper and sixth or seventh lower largest, fourth or fourth and fifth upper reaching eye, first pair of lower meeting on median line. Genials in two pairs, posterior a little longer, anterior touching five labials. Scales on body smooth, without pits, in nineteen rows. Anal plate divided. Gas- trosteges two hundred and nine. Tail complete. Urosteges one hundred and eight, all paired. The color above is a pale grayish olive. A dark streak runs back from the eye. The infralabials and the posterior superior labials are blotched with yellowish white. There is a yellowish-white blotch on each side of the nape. There are no traces of longitudinal bands on the body, but along the back is a series of irregular dark brown cross bars or alternating spots. A few indications of similar spots may be made out on the sides. The tail is unspotted except near its base. The lower surfaces are creamy white, plentifully dotted or clouded with dark gray. There are no very distinct blackish-brown lateral spots on the anterior gastrosteges. Length to anus, 590 mm. Length of tail, 220 mm. Variation—The type specimen from Charles Island has two hundred and nine gastrosteges, one hundred and ten uros- 338 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. teges, three postoculars, scales in nineteen rows, and the spot- ted style of coloration. General Remarks.—Snakes must be very rate on Charles Island, for none were seen there by any member of our expe- dition, although careful search was made for them. It is probable that the ravages of the smaller kinds of mammals that have been introduced there—particularly rats and cats— have pushed them to the verge of extinction, as they have the Tropidurus. It is probable that a longer search would show that snakes are still to be found on Champion and Enderby as well as on Gardner, for Tropiduri still are fairly abundant on all these islets. The Charles Island snake is most closely related to the Hood Island species. It differs from that species in having numerous dorsal spots, no dorsolateral bands, and no definite dark spots on the anterior gastrosteges. Dromicus hoodensis new species. Hoop ISLAND SNAKE 1892, Orophis biserialis GARMAN, Bull. Essex Inst., XXIV, 1892, p. 85 (part). 1903, Dromicus biserialis habeli HELLER, Proc. Washington Acad. Sci., V, 1903, p. 93. Diagnosis.—No scale-pits ; scales in 17 or 19 rows; gastros- teges 203 to 214; urosteges 91 to 114, usually all paired ; post- oculars three; temporals usually 2+2; never spotted; striped, the stripes becoming obsolete posteriorly. Type-——Male. California Academy of Sciences No. 11799. Hood Island, Galapagos Archipelago. J. R. Slevin. June 23, 1906. Distribution—Hood and Gardner-near-Hood islands, Gala- pagos Archipelago. Material—One specimen collected by Dr. Baur on Hood Island has been recorded by Garman. Two secured on Hood by Heller are Nos. 4970 and 4971 in the collection of Stanford University. The Academy has thirty-four from Hood and one from Gardner-near-Hood. Vor. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 339 Description of the type-—Head rather long, with flattened top and rounded snout. Rostral plate large, a little broader than high, hollowed below, and bounded behind by internasal, anterior nasal and first labial plates. Plates on top of head are: a pair of internasals, a pair of prefrontals, supra- ocular and part of preocular of each side, a frontal, and a pair of large parietals. Internasals much smaller than prefrontals. Frontal slightly longer than parietal suture. Anterior and posterior nasals distinct. Loreal well developed, longer than high. One preocular. Postoculars three. Temporals two followed by two. Eight superior and nine inferior labials, sixth upper and fifth or sixth lower largest, fourth and fifth upper reaching eye, first pair of lower meeting on median line. Genials in two pairs, posterior a little longer, anterior touching five labials. Scales on body smooth, without pits, in seventeen rows. Anal plate divided. Gastrosteges two hundred and seven. Tail complete. Urosteges one hundred and thir- teen, all paired. The color above is deep olive brown becoming paler posteriorly and seal brown toward the head. A light dorsolateral band, about two scales wide, arises on the upper postocular, crosses the parietal, and continues along the fifth and sixth rows of scales. This yellowish-brown band becomes less distinct on the middle third of the body and nearly obsolete posteriorly. The tail is unicolor, olive, becoming yellowish olive toward the tip. The dark brown postocular or temporal bar is continuous with the brown band on the side of the neck. There is no light nuchal blotch. The labials are yellowish white marked with blackish olive. The first and second rows of scales on the neck are whitish, marked anteriorly with a row of blackish spots, continuous with a similar row formed of one spot near the lateral extremity of each gastrostege from about the fourth to twenty-second. The lower surfaces are yellowish dotted or clouded with grayish olive. Length to anus, 518 mm. Length of tail, 217 mm. Variation.—All the males have seventeen rows of scales, while all of the females have nineteen rows. Careful search of every specimen failed to disclose any scale-pits except in the case of No. 9306, on which one scale showing a single pit was found. The gastrosteges range in number from two hundred and three to two hundred and fourteen. The uros- teges in specimens with complete tails vary from ninety-four to one hundred and fourteen in males and from ninety-one to one hundred in females. All of the urosteges are paired except in the specimen from Gardner Island, which has two undi- vided. The postoculars are always three. The temporals normally are 2+-2, and the supralabials eight. The following table shows the scale-counts. In the urostege column c¢ indi- cates that the tail is complete, while + is affixed to counts when the tip of the tail is missing. January 15, 1912 340 “CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. TABLE OF SCALE COUNTS, Dromicus hoodensis a oo n & n Fs & B | s g 3 | ¢ bs a ie £ 3 5 8 & Ag) = E * 7” Gg 3 8 2 & S a) 3 8 s 3 & u fe} 3) & 3 Zz a}|a|]o Bo] a | a a ae 9336 ry 17 | 206 |105 c 1 3 | 2+2 8 il 9370 $ 17 |211|109 c 1 3 | 2+2 8 1 9420 é 17 | 208 |110 c 1 3 | 2+2 8 1 10920 é 17 |204|112 c il 3 | 2-2 8 1 10957 é 17 | 214/113 c 1 3 | 2+2 8 1 11799 3} 17 | 207 |113 c 1 3 | 2+2 8 1 | Type 11800 6 | 17 |206|107 c 1 3 | 2+2 8 1 11896 3 17 |208; 68 +} 1 3 | 2+2 8 1 11921 re} 17 | 208 |106 c 1 3 | 2+2 8 1 11923 3 17 | 205105 c 1 3 | 2+2 8 1 11931 re) 17 | 207 |108 c 1 3 | 242 8 1 11932 6 | 17 |207| 94 c¢c 1 3 | 2+2 8 1 11934 3 17 | 209 }112 c 1 3 | 2+2 8 1 11936 3} 17 | 208)114 c 1 3 | 2+2 8 1 11937 3 17 |205| 33 +) 1 3 | 2+2 8 1 11939 3 17 |208|107 +} 1 3 | 2+2 8 1 | Gardner Island 9304 @ | 19 |...) 84 +) 1 3 | 2+2 8 1 9305 2 | 19 }205} 93 c 1 3 | 2+2 8 1 9306 @ | 19 }208) 89 +) 1 3 | 24+2 8 1 9335 @ | 19 |210] 96 +] 1 3 | 2+2 8-9] 1 9384 @ | 19 |210} 80 +] 1 3 | 2+2 8 1 10919 2 | 19 }210} 90 +] 1 3 | 2+2 8 1 10921 @ | 19 |}207| 92 +) 1 3) |) 24-2 8 1 10922 2 | 19 |206] 94 c 1 3 | 2+2 8 i] 10958 @ | 19 }208| 90 +] 1 3 | 2+2 8 1 11895 @ | 19 |}211} 96c 1 3 | 2+2 8 1 11920 Q@ | 19 1207] 95 ¢ 1 3 | 2+2 8 sl 11922 @ | 19 ,211) 95 +] 1 3} 2+2 8 1 11924 Q | 19 | 204] 95 c¢ 1 3 | 2+2 8 1 11925 2 | 19 |209} 96c 1 3 | 2+2 8 1 11926 @ | 19 |205} 93 ¢ 1 3 | 2+2 8 1 11930 @ | 19 |207|] 91c¢ 1 3 | 2+2-1+2) 9 1 11933 2 | 19 | 203] 94 c¢ 1 3 | 242 8 1 11935 9 | 19 }210} 90 +} 1 3 | 242 8 1 11938 @ | 19 | 209 |100 c 1 3 | 242 8 1 Garman 100209 ee ee 1 SES coe 8 1 S.4970 @ |} 19/212) 94 +) 1 3 | 2+2-2+4-2) 8 1 | Stanford Univ. S.4971 $ | 17 |206|] 98 c 1 Sheet, 8 1 | Stanford Univ. Brit. Mus.| 6 203 |104 Biya) eer tears me a Brit. Mus.| 6 199 }105 Brit. Mus.| 2 7A S50 Boe There is very little variation in coloration. All specimens are striped without trace of dorsal spots; and in all, the stripes fade out posteriorly. In specimens with nineteen rows of scales the stripes are on the sixth and seventh rows. All specimens but one show the characteristic spotting on the anterior gastrosteges only, with the white continuation of the Vor. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 341 labial streak just above. In the one exception, a young speci- men, there are mere traces of the dark spots. The largest specimen measures 820 mm. from snout to vent, and has a tail 253 mm. long. Habits —Nothing is known of the breeding habits of any of the Galapagos snakes. One of the Hood Island specimens (No. 9306) contained the tail of a large Tropidurus which it had eaten. General remarks.—Snakes still are abundant on Hood Island. They seem to differ from those of Charles Island only in coloration; but, since the differences are constant in the large series at hand, they must be regarded as a distinct species. The sexual difference in the number of scale rows in the snakes of this one island is worthy of note. Dromicus dorsalis (Steindachner). GaLapacos SNAKE 1869, Dromicus chamissonis Peters, Mon. Berlin. Acad., 1869, p. 719; Gtntuer, Zool. Record, 1869, p. 115 (part) ; BouLtENcErR, Cat. Snakes Brit. Mus., II, 1894, p. 119 (part). 1876, Dromicus chamissonis var. dorsalis STEINDACHNER, Festschr. Zool.-bot. Ges. Wien, 1876, p. 306, pl. I, fig. 1 (type localitiesIndefatigable [probably] or Jervis islands). 1876, Dromicus chamissonis var. Habelii STEINDACHNER, Festschr. Zool.- bot. Ges. Wien, 1876, p. 306, pl. I, fig. 1 (type localities Indefatigable [probably] or Jervis islands). 1889, Opheomorphus chamissonis Corr, Proc. U. S. Nat. Mus., XII, 1889, p. 147. 1892, Orophis biserialis GARMAN, Bull. Essex Inst., XXIV, 1892, p. 85 (part). 1903, Dromicus biserialis biserialis HELLER, Proc. Washington Acad. Sci., V, 1903, p. 93 (part). Diagnosis.—Scale-pits present; scales in 19 rows; gastros- teges 213 to 236; urosteges 95 to 119, usually some unpaired ; postoculars two, rarely one; temporals usually 1+2 or 1+1; usually striped, sometimes spotted (on Barrington and Inde- fatigable). Types.—Vienna Museum. Galapagos Archipelago, prob- ably Indefatigable (or Jervis). Dr. Habel. 1868. 342 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. Distribution James, Jervis, Indefatigable, South Seymour and Barrington Islands, Galapagos Archipelago. Material.—Five specimens collected by Dr. Habel, probably on Indefatigable or Jervis Islands, are in the Vienna Museum. I have examined fifty-one specimens in the Academy collec- tion, as follows: twenty-three from Indefatigable, fifteen from Barrington, eight from James, three from South Seymour, and two from Jervis. Description of No. 12062—Adult male. Indefatigable Island. J. R. Slevin. July 16, 1906. Head fairly broad, with flattened top and rounded snout. Rostral plate large, much broader than high, hollowed below, and bounded behind by internasal, anterior nasal and first labial plates. Plates on top of head are: a pair of internasals, a pair of prefrontals, supraocular and part of preocular of each side, a frontal, and a pair of large parietals. Internasals smaller than prefontals. Frontal longer than parietal suture. Anterior and posterior nasals distinct. Loreal well developed, little longer than high. One preocular. Two postoculars. Temporals one followed by two. Eight superior and ten inferior labials, sixth upper and fifth or sixth lower largest, fourth and fifth upper reaching eye, first pair of lower meeting on median line. Genials in two pairs, posterior a little longer, anterior touch- ing five labials. Scales on body smooth, many with pits, in nineteen rows. Anal plate divided. Gastrosteges two hundred and twenty. Tail complete. Urosteges one hundred and thirteen, the second to seventh undivided. The upper surface of the head is yellowish olive dotted with brown. There is a brown band from the rostral to the eye and from the eye to the side of the neck. The labials, chin, and throat are yellowish white dotted with dark brown. The body is longitudinally striped. The lower three (or, on the posterior part of the body, two) rows of scales are gray- ish brown. The next row is dark brown. The fifth, sixth, and seventh rows are yellowish white. The eighth row is dark brown, and the three rows along the middle of the back are lighter olive brown. The stripes are continued on to the tail, but the distal portion of this region is plain yellowish olive. The lower surfaces are yellowish white irregularly dotted and spotted with dark brown. Length to anus, 670 mm. Length of tail, 248 mm. Variation: Indefatigable Island.—Only two specimens (Nos. 10233 and 10796) have all the urosteges divided. Some specimens have only the second urostege undivided. At the other extreme is No. 10232 in which the second to twenty- second, forty-fifth to forty-eighth, and sixty-third to sixty- fifth, are unpaired. The urosteges range from one hundred and five to one hundred and nineteen, and the gastrosteges from two hundred and seventeen to two hundred and thirty. Vor. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 343 TABLE OF SCALE COUNTS, Dromicus dorsalis (Steindachner) INDEFATIGABLE ISLAND . E g g CN RE: HN (mle a hae 2 = ¢ 8 ells = aiid rm cS ra 3 8 % 5 a |e z a B é 5 caitn (tee es a a |a 10232 | ¢ 19 | 224 | 118 c LA PS TTT | UOT gi 10234 | 8 19), 1222) Mitac Ui (Ae ee POE I al 10235 | 6 19 | 226 Ose Eee tny||y Qed MiMysinied 10303 | 3 19 | 219 GG) Se MN VAT aa realty |i esl 10304 | 6 19 | 224 COMPS Weedeo st 26 ll teed sy bese et 10305 | 3 LOM 2o Te OT OSPeEN nah O) eeeo i eels 103750 18 19 | 225 | 114c Tost (noel a ae SS ates 8 as) 10378 | 8 19), 225) e107) c floed lari 2g le Si be 10379 | 8 19) 419 2234) itz Ae hse |e dete ee om ied 10395 | 8 19 | 226 | 119 c Tp eel here tn hi Sb 10396 | 6 TESA Mi | PS MTP RR 27 a 10559 | 6 OME TaM ee O Nye site ale Du lle deen iN seielied 12056 | 8 19 | 224 | 116c HE Deni maaan |i eel al 12059 | 3 19 | 229 | 105c yas) dete oe ee ea 12062 | 8 19} 220) Mh dik3 rc Tha lie) arte ASE 12063 | 6 Tye) Baar LO ete a ge ee aa ate 12064 | ¢ OY PSII Ss eet si (Py Jaa eer api rcs 12065 | 3 LON DIS MOTOR May On| gato Neg ny iid 1023315 | Qt 70) 228 HOME Pe ninth | 2s |) Sait D8 ols Sheela 10429 | @ | 19 | 230 | 107 c¢ 1a oA 2b | deat 2 eral San 1056009], 9 |\19-7)) 225 Spey ei leteals|).( 2). 1todh,|8 Sauls fO79D>) 9) 49) 1|\ 299 Tae paleo 2h | dee ole MeaN ed 110796) ll) 09: I) 19)!11.220) 5405 c (hacen rake es ona (een The largest snake in the collection is No. 10792 which measures 950 mm. from snout to anus. All but six of the Indefatigable snakes are colored like the one described above. Seventy-four per cent of the specimens from this island are striped. Of the remaining six specimens, two (Nos. 10233 and 12064) are spotted to the tail, while the other four (Nos. 10235, 10305, 10379, and 10792) are spotted anteriorly, but become nearly unicolor, or at most show only faint spots and bands posteriorly. Nos. 10233, 12064, 10235, and 10305 show longitudinal light stripes more or less clearly” on the posterior part of the body. These stripes are wanting in Nos. 10379 and 10792. The light stripes or nuchal blotches are continued forward very distinctly to the parietals in all Indefatigable specimens except 10235, 10305, 10379, 10396, and 10792. The light ’ stripes, when present, never are confined to the scales of two rows, as is the case in snakes from Barrington. South Seymour Island——Two of the three specimens at hand have more numerous gastrosteges than have been found 344 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. in any specimen from Indefatigable, James, Jervis, or Bar- rington islands. The third has a number equaled by only one Indefatigable specimen. In other respects these snakes are like the Indefatigable striped specimens, except that the color- ing is a little lighter and brighter. TABLE OF SCALE COUNTS, Dromicus dorsalis (Steindachner) SOUTH SEYMOUR ISLAND 3 e g 2 Fa 3 eelae a a oO oO p [a0 a a a ya 10483 é 19 232 88 + 1 wy) 1-42 8 1 10485 3é 19 230 113itc 1 2 1+2 8 1 10486 2 19 Xie | coe 1 2-3 2 8 1 James Island.—The James Island snakes show no important differences from the Indefatigable series. Nos. 10782 and 12153 have all urosteges paired. No. 12091 has the second to twenty-first undivided. No. 12092 has a similar condition in the sixth to eighth, tenth to fifteenth, seventeenth, nineteenth, twenty-first to twenty-third, twenty-seventh, and one-hun- dredth to one-hundred-and-third. All the others have some unpaired. The temporals usually are one followed by one. Variation in other scale-characters is shown in the following table: TABLE OF SCALE COUNTS, Dromicus dorsalis (Steindachner) JAMES ISLAND a® % n fe! E 24 if 5 | 88 3 = ch |b & 5 cel g 20 2 g 3 3 8 £ $ 3 * a a & 2 fo} 3 & 8 Z a a o =) a [4 a a 35) 12091 | ¢ | 19 | 217 Of ae i aed alee lea 12092 | @ 19 | gis) tod | ee {ie 8 | 4 12004 | @ | 19 | 213 G8i2 || oie Wena al |i tbennl ee ale 12154 | 3 | 19 | 213 72 et eitia saat 8 | 4 12155 | ¢ | 19 | 213 | 101c 10 Weoaaleet eed als Ihe 10782 | @ | 19 | 226 Taste atone HES 8 | 4 120933 | 9 | 19 | 221 85 Ete alee A g |1 12153 | @ | 19 | 220 yiweeat) |hoe] Ts 8 |1 Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 345 All the specimens are striped. The stripes are clear and distinct except in Nos. 12093, 12094 and 12154, in which they are more or less obsolete behind the neck. They are continued to the parietals, and usually involve the scales of three rows. Jervis Island——Two specimens from Jervis seem to agree perfectly in squamation and coloring with the James Island snakes. -TABLE OF SCALE COUNTS, Dromicus dorsalis (Steindachner) JERVIS ISLAND E 2 Bi g g | & 3 Z Z e | 8 Seen Be | wae Mbabllcs 3 % 3 8 3 | a 2 z a |e is 5 alee) Ny es | ae Iss OGIO eae) to) i220) eos ed 2 14 tts, Ns a - 10611 g 19 226 Slate 1 2 ise 8 1 Barrington Island.—l have before me fifteen snakes from Barrington. All but four of these have a few urosteges un- divided. The .ariation in important scale-characters is set forth below. The tendency toward a reduction in the number of temporals and postoculars will be noted. TABLE OF SCALE COUNTS, Dromicus dorsalis (Steindachner) BARRINGTON ISLAND 3 Z 8 8 ge | & 3 = a 5 g By soll 8 2 fala 5 ¥ 3 a 3 8 | 2 & Sande Zz a a co) 5 a a a came 10147 | @ | 19 | 223 86 ie Maha ae deed eae os hs 10152 | 2 19 | 220 95 c F109 ea pa eee ea ee 10182 | @ 19 | 218 Ait Ote a O|. tees pest at 10183 | 3 19 | 218 9g ena a! | caer erSe ed 10215 | @ 19 | 218 Est 4 No an Ra ied (are Ui 10217 | @ 19 | 222 | 104c ri | Sa ieee et Wd eso 10226 | @ 19 | 219 eiffel] 10, | ee peepee A (| 12061 | @ 191) |) 205 Bc) dee ee? eis 8 |1 10150 | @ | 19 | 207 St cy etalbt=9 | deve esis 10151 | @ | 19 | 229 SG ute on esmeae ed oi:|) ea lige ira 10213 | ° | 19 | 223 ryote | Ae), ||| RO amr Ege ROU 10214 | 2 | 19 | 229 Gi n=: |) 123 bese a 10216 | @ | 19 | 226 6S RA ame! | een Pcgh 4 12055 | 9 | 19 | 223 eye || A) |] oem Pmoeeor ried Pests 12060 | 9 | 19 | 227 ieee Nf. it || IN Ret cl 2 Brit. Mus.| ¢ 224 ada fels we tb ee me “fs 346 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. Two styles of coloration are exhibited by the snakes of Barrington. Seven specimens (Nos. 10151, 10152, 10183, 10213, 10214, 10217, 10226) are striped, while eight (Nos. 10147, 10150, 10182, 10215, 10216, 12055, 12060, 12061) are spotted. The difference here, as on Indefatigable, is due neither to age nor sex. It must be regarded as a form of dichromatism. In the spotted specimens, the longitudinal light stripes are represented only by a pair of short longitudinal yellowish-white blotches on the nape. In striped specimens, the light stripes are confined to the scales of two rows. In all specimens, the light nuchal blotches or the longitudinal stripes end anteriorly sharply and definitely several scales behind the parietals. In all spotted specimens, the dark brown spots or blotches become obsolete posteriorly ; while, in all striped speci- mens, the light bands extend to the tail. General Remarks.—It is probable that larger series may result in the recognition of subspecies of Dromicus dorsalis. Even now the peculiarities of coloration, with the frequent reduction in temporals and postoculars, almost justify the separation of the Barrington Island snakes. The serpents of Indefatigable and Seymour appear to differ from those of the other islands in the possession of a greater number of uros- teges, but so many of the specimens have lost the tips of their tails that more evidence is needed. Inconstant as the differ- ences may prove to be, I believe that the following tentative key may prove useful to future investigators. a.—Stripes or nuchal blotches ending definitely several scales behind par- ietals; stripes narrow. Temporals usually 1+1; postoculars often 1; urosteges fewer. Barrington. a.2—Stripes or nuchal blotches usually continued forward to parietals; stripes wider. b.—Urosteges more numerous; temporals usually 1+2; spotted or striped. Indefatigable and Seymour. b.2—Urosteges fewer; temporals usually 1+1; striped. James and Jervis. We do not know why so many of these snakes have lost the tips of their tails, but Mr. Slevin reports that the mocking- birds were observed picking at the tails of Tropiduri until they fell off and could be eaten. . Vot. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 347 Dromicus occidentalis, new species. NARBoROUGH ISLAND SNAKE 1903, Dromicus biserialis biserialis HELLER, Proc. Washington Acad. Sci., V, 1903, p. 93 (part). Diagnosis.—Scale-pits present; scales in 19 rows; gastros- teges 236 to 252; postoculars two; temporals 1+1 or 1+2; striped (or rarely spotted), light nuchal blotches and series of dark spots on tips of gastrosteges and on lower lateral scales very distinct. Type.—Adult female. California Academy of Sciences No. 11488. Narborough Island, Galapagos Archipelago. J. R. Slevin. April 18, 1906. Distribution.—Narborough Island, Galapagos Archipelago. Material—Mr. Heller has recorded four snakes from Nar- borough, now forming a part of the collection of Stanford Uni- versity, where I have examined them. The Academy has received only two from Narborough. Description of the type—Head rather broad, with flattened top and rounded snout. Rostral plate large, much broader than high, hollowed below, and bounded behind by internasal, anterior nasal, and first labial plates. Plates on top of head are: a pair of internasals, a pair of pre- frontals, supraocular and part of preocular of each side, a frontal, and a pair of large parietals. Internasals much smaller than prefrontals. Frontal longer than parietal suture. Anterior and posterior nasal distinct. Loreal well developed, longer than high. One preocular. Two postoculars. Temporals one followed by two, or one followed by one. Eight superior and ten inferior labials, sixth upper and fifth or sixth lower largest, fourth and fifth upper reaching eye, first pair of lower meeting on median line. Genials in two pairs, posterior a little longer, anterior touching four labials. Scales on body smooth, many with pits, in nineteen rows. Anal plate divided. Gastrosteges two hundred and forty-seven. Tail incom- plete. Urosteges ninety-eight, all paired. The top of the head is dark brown mottled with olive gray. A light brown band extends from the rostral plate to the eye, and a dark brown postocular blotch crosses the temporal region to the side of the neck. The labials and lower surfaces of the head and throat are olive gray marbled with dark brown. On each side of the body there is a light yellowish-gray longitudinal stripe along the sixth and seventh rows of scales. On the posterior portion of the body, where there are only seventeen rows of scales, this stripe drops to the fifth and sixth rows. It is continued beyond the middle of the tail; but on the neck, as far as the twenty-fifth gastros- tege, it is represented by a series of nine large, rounded, light spots. Along the back between these light stripes is a band of dark brown, darker on the scales bordering the light stripes. The sides are dark brown close to the lateral light stripes, but become grayish olive toward the gastrosteges. On the anterior half of the body, most of the scales of the second row, and a few of those of the first, bear central spots of dark brown. Similar small 348 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. blackish-brown spots on the tip of each gastrostege form a row extending nearly to the tail. The lower surfaces are yellowish with numerous small blackish spots. Length to anus, 890 mm. Length of tail, 252 mm. Variation.—No. 11509 has the first eleven urosteges undi- vided. These scales are all paired in all of the other specimens except No. 4974 of the Stanford University collection, in which the first urostege is unpaired. The following table shows the principal variation in squamation. TABLE OF SCALE COUNTS, Dromicus occidentalis, new species. Number Scale rows Gastrosteges Urosteges Preoculars Postoculars Temporals Supralabials Loreal Sex 11509 19 | 237 Bone 2 |iitt = 11488 8.4974 8.4975 8.4973 8.4976 19 247 98 + 116 c 19 236 112 19 252 91 + 19 243 109 c 40 10 0303 49 3 a oO bdo w oO ee eee wo NY NH tH — Ne Cc co MOCO oO lo) All the Narborough specimens have the characteristic light nuchal blotches and dark spots on gastrosteges and lower lateral scales. The dark spots on the lower laterals are most numerous on the scales of the first row in all specimens except the type. The row of spots along the tips of the gastrosteges extends to the vent in No. S. 4975 and S. 4976, nearly to the vent in No. S. 4974, and past the middle of the body in No. S. 4973. All the specimens show the longitudinal light stripes except No. S. 4975, which is spotted without any trace of stripes. The general dorsal coloration of this specimen is similar to that of the snakes of Albemarle and Brattle, but it shows the light blotches on the nape, and dark spots on gastrosteges and laterals, which are characteristic of the Narborough snakes. No. 11509 is intermediate in coloration between No. S. 4975 and the other Narborough specimens. It shows both stripes and spots most distinctly. General remarks.—The snakes of Narborough agree with those of Albemarle and Brattle in the large number of their Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 349 gastrosteges, a character which distinguishes them from all other Galapagos snakes. They seem to differ from those of Albemarle and Brattle only in coloration; and, since two speci- mens show a tendency to vary in the direction of the Albe- marle form, it seems best to regard those from Albemarle as a subspecies. Dromicus occidentalis helleri, new subspecies. HELLER’S GALAPAGOS SNAKE 1903, Dromicus biserialis biserialis HELLER, Proc. Washington Acad. Sci., V, 1903, p. 93 (part). Diagnosis.—Scale-pits present, scales in 19 rows; gastros- teges more than 236; postoculars two; temporals 1-+-2 or 2+2, spotted, no longitudinal light stripes; no series of definite rounded blackish spots on lateral scales of first and second rows; light nuchal markings much less prominent, and dark spots on tips of gastrosteges absent or less distinct than in the Narborough form. Type.—Male. California Academy of Sciences No. 10280. Brattle Island, Galapagos Archipelago. J. R. Slevin. October 30, 1905. Distribution.—Albemarle and Brattle Islands, Galapagos Archipelago. Material.—Mr. Heller has recorded one specimen from near Cape Berkeley, Albemarle, which now is No. 4977 of the Stanford University collection. The Academy has received two from Brattle. Description of the type-—Head rather broad, with flattened top and rounded snout. Rostral plate large, much broader than high, hollowed below, and bounded behind by internasal, anterior nasal, and first labial plates. Plates on top of head are: a pair of internasals, a pair of pre- frontals, supraocular and part of preocular of each side, a frontal, and a pair of large parietals. Internasals smaller than prefrontals. Frontal longer than parietal suture. Anterior and posterior nasals distinct. Loreal well developed, little longer than high. One preocular. Two postoculars. Temporals one followed by two. Eight superior and ten inferior labials, sixth upper and fifth lower largest, fourth and fifth upper reaching eye, first pair of lower meeting on median line. Genials in two pairs, posterior longer, anterior touching four or five labials. Scales on body smooth, many with pits, in nineteen rows. Anal plate divided. Gastrosteges two hundred and forty. Tail complete. Urosteges one hundred and twelve, the first to third, seventh to eleventh, and fourteenth and fifteenth not divided. The top of the head is olive brown dotted with olive gray. A light brown band extends from the rostral plate to the eye, and a brown post- 350 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. ocular blotch crosses the temporal region to the side of the neck. The labials and lower surfaces of the head and throat are yellowish-gray marbled with dark gray. There are no light longitudinal stripes on the body. The color above shades from brownish olive along the middle of the back to pale olive gray near the gastrosteges. On the neck are large round dark brown spots separated by light yellowish-gray blotches. On the anterior part of the body these dark spots become smaller and more numerous, and form three alternating rows on each side. These spots become smaller and less numerous posteriorly, and are lacking on the tail. They also tend to avoid the sixth and seventh rows of lateral scales. The lower surfaces are yellowish mottled with brownish gray except on the tail. Many of the tips of the gastrosteges bear not very definite small dark brown spots, but there is no series of such spots on the lower lateral scales. Length to anus, 542 mm. Length of tail, 178 mm. Variation.—The Albemarle specimen has the upper post- ocular of one side united with the parietal. It has eight supe- rior and ten inferior labials, the fourth and fifth upper reaching eye, the sixth in each series largest, five inferior in contact with the anterior genial. Both it and No. 10281, from Brattle, have all urosteges divided. TABLE OF SCALE COUNTS, Dromicus occidentalis helleri, new sub-species ill ue g Polya 5 5 2 . g| 3 5 3 r a A) 2 a] 8 a [ia Be SUM re a ee Mi Fee eh GPEC Zz, a a} oO rs) eg |) By a AIA 4 10280 | @ |19|240/112c |1| 2 | 142 |8| 1 | Brattle 10281 | 9 | 19] 248] 98 +]1]| 2 | 142 |8| 1 | Brattle 54977 | 9 | 19| 241] 88 +11] 1-2 (ce 8 | 4 | Albemarle The two Brattle snakes are absolutely alike in coloration, and the Albemarle specimen is very similar, as will be seen from the following description of Stanford University No. 4977, adult female, from vic. Cape Berkeley, Albemarle Island. The head :s brownish olive marbled with black. There is a dark post- ocular or temporal streak. The labials are mottled with lighter. There are no longitudinal light lines. The upper surfaces are dark brown spotted with darker brown or black. On the neck, these spots are large, round and very distinct and well defined. On the body, they are smaller and become perhaps less distinct toward the tail. Still, they form, throughout the whole length of the body, two alternating rows usually on the fifth and eighth rows of scales of each side, dropping to the fourth and seventh rows posteriorly. There are two or three pairs of whitish blotches on the nape. The lower surfaces are yellowish irregularly spotted with brownish black. Almost every gastrostege on the anterior two-thirds of the body shows a definite blackish spot near its outer extremity on each side as in oe Narborough snakes, but there are no similar spots on the first row of scales. 2 Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 351 General remarks.—I take pleasure in naming this snake after Mr. Edmund Heller who collected the Albemarle specimen while a member of the Hopkins-Stanford Galapagos Expedi- tion in 1898-99. Dr. Boulenger writes me that the British Museum has a young spotted snake said to have been collected at Tagus Cove, Albemarle. It has one hundred and twelve urosteges, but only two hundred and twenty-two gastrosteges. This small number of gastrosteges makes me think that an error may have been made in the locality label. The specimen has scales with two pits, and one would incline to the opinion that it has originated on Barrington or Indefatigable. If, however, there has been no mistake in the label, the Tagus Cove snakes must represent a species distinct from that found at Banks Bay; and it may be that larger collections will show that each of the five large mountains of Albemarle has its own peculiar race of serpent. Dromicus slevini, new species. SLEVIN’s SNAKE 1903, Dromicus biserialis biserialis, HELLER, Proc. Washington Acad. Sci., V, 1903, p. 93 (part). Diagnosis—No scale-pits; scales in 19 rows; gastrosteges 170 to 183; urosteges 82 to 104; no longitudinal light stripes. Type—Male. California Academy of Sciences No. 12,216. Duncan Island, Galapagos Archipelago. August 14, 1906. Distribution—Duncan, Narborough, and Cowley Moun- tain, Albemarle. Material—tThree specimens are known. Two are in the Academy collection, while the one from Narborough belongs to Stanford University. Description of the type-—Head rather broad, with flattened top and rounded snout. Rostral plate large, broader than high, hollowed below, and bounded behind by internasal, anterior nasal, and first labial plates. Plates on top of head are: a pair of internasals, a pair of prefrontals, supraocular and part of preocular of each side, a frontal, and a pair of large parietals. Internasals much smaller than prefrontals. Frontal slightly shorter than parietal suture. Anterior and posterior nasals distinct. Loreal well developed, longer than high. One preocular. Two post- oculars. Temporals two followed by two, or one followed by one. Eight superior and ten inferior labials, sixth upper and sixth lower largest, fourth and fifth upper reaching eye, first pair of lower meeting on median line. Genials in two pairs, posterior a little longer, anterior touching five labials. Scales on body smooth, without pits, in nineteen rows. Anal plate 352 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. divided. Gastrosteges one hundred and eighty-three. Tail complete. Urosteges one hundred and four, all paired except the first to fourth. The head is brownish olive above, with whitish spots on the labials and a dark brown postocular streak. The back is crossed by about fifty-five black cross-bars separated by narrower brownish-white ones. In some places the black bars are not quite continuous, tending to alternate at the mid-dorsal line with those of the opposite side of the body. These black cross-bars extend down on the sides to about the second row of scales. The other lateral scales are of a brownish-gray color, continuous with the light cross-bars, and are sometimes outlined with black. The tail is pro- vided with about thirty blackish-brown blotches proximally, becoming uni- color toward the tip where it is olive. The lower surfaces are grayish, more or less dotted with slate, and the base of each gastrostege shows a more or less concealed blackish cross-bar. Length to anus, 228 mm. Length of tail, 95 mm. Variation.—The principal variation in scale characters is set forth in the following table. TABLE OF SCALE COUNTS, Dromicus slevini, new species 2 & 8 g 2 ic a eee a lees dies 3 e| 3 Salis ays Ee z § : a] 4 4 3 | 3 g E| 2 g 3 oO oO s =] i {o) oO i} co) ° Zi n n Lo) p [a¥) eM] & aly 4 12216 @ | 19/183 | 104c] 1] 2-2 ae 1 | Duncan 121501 9 |191170| 82cl 1| 22] 242 | 8/1 | CowleyMt. 5.4972} 9 119/179] 96cl| 1] 2-2 as 8 | 1 | Narborough In all the specimens except the type all of the urosteges are divided, and the frontal is slightly longer than the parietal suture. Neither the Duncan nor the Cowley Mountain speci- men shows any trace of longitudinal light stripes. Both are, in general, black with vertical light bars on the sides. In the Duncan snake most of these light bars cross the back; while in the Cowley specimen they do not extend above the lateral regions, leaving a black dorsal band three or four scales wide. The Narborough specimen agrees in coloration with that from Cowley Mountain. The Cowley specimen has about eighty- five light bars on the upper part of each side, where the Nar- borough snake has only seventy-one, and the Duncan about fifty-five. In the Cowley and Narborough snakes these light bars fork inferiorly and, joining with branches of the preceding Vor. I] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 353 and succeeding bars, outline alternating dark spots on the lower lateral scales. The largest specimen of D. slevini is that from Cowley Mt., Albemarle Island, which measures 347 mm. from snout to anus, and 135 mm. from anus to tip of tail. Habits—The Duncan Island snake contains the foot and tail of a gecko which it had eaten. General Remarks.—The Cowley Mountain snake was taken August 11, 1906, on a field of pumice stone at an elevation of about 200 feet. Mr. Slevin’s notes state that it was the only snake secured on Albemarle, and differed in coloration from any taken elsewhere. Under date of August 14, 1906, Mr. Slevin wrote: ‘“Anch- ored off Duncan about ten a.m. I collected on the northeast slope of the island to about 800 feet. Got a snake at about 400 feet. It appeared different from any taken thus far.. It was very well colored to prevent detection. It was secured on a lava block covered with silver colored lichen which matched the snake exactly. One was reported by Mr. Hunter during our last stop at Duncan, which, he said, was similar in coloring to the one taken today.” Mr. Drowne of the Webster-Harris Expedition reports* having seen on Duncan Island, September 9, 1897, a snake that was about one and a half feet long, slender and blackish, with white rings. It is probable that more abundant material will show that more than one species has been included here under the name Dromicus slevini. Dromicus steindachneri, new species. STEINDACHNER’S SNAKE Diagnosis——No scale-pits; scales in 19 rows; gastrosteges 169 to 180; urosteges 96 to 114; longitudinal light stripes present. Type.—Male. California Academy of Sciences No. 10795. Indefatigable Island, Galapagos Archipelago. J. R. Slevin. Jan. 16, 1906. Distribution.—This species has been found on Indefatigable, South Seymour and Jervis islands. It is probable that more Minoninrcs Zooks VI" po! 117, 354 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. extensive collecting will show that it is present also on James and Barrington. Material—We have received five specimens. Two are from Jervis, two from South Seymour and one from Indefatigable. Description of the type—Head rather broad, with flattened top and rounded snout. Rostral plate large, broader than high, hollowed below, and bounded behind by internasal, anterior nasal, and first labial plates. Plates on top of head are: a pair of internasals, a pair of prefrontals, supraocular and part of preocular of each side, a frontal, and a pair of large parietals. Internasals much smaller than prefrontals. Frontal slightly longer than parietal suture. Anterior and posterior nasals distinct. Loreal well developed, longer than high. One preocular. Two postoculars. Tem- pora!s two followed by two, or one followed by one. Eight superior and ten inferior labials, sixth upper and sixth lower largest, fourth and fifth upper reaching eye, first pair of lower meeting on median line. Genials in two pairs, posterior a little longer, anterior touching five labials. Scales on body smooth, without pits, in nineteen rows. Anal plate divided. Gastrosteges one hundred and sixty-nine. Tail complete. Urosteges ninety-six, all paired. The head is brownish olive above. There is a dark brown postocular streak. The labials and most of the other scales on the side of the head are yellowish gray with dark borders. The general color above is blackish brown. A light yellowish-gray stripe runs along each side of the neck, body, and tail. This streak is on the scales of the sixth, seventh, and eighth rows on the neck, and of the fifth, sixth, seventh and sometimes eighth on the body, except posteriorly where it drops to the fourth, fifth, and sixth rows. Many of the lateral scales have light central spots of the same color as the longitudinal stripes. The lower surfaces are light yellowish gray. There is a blackish cross-bar at the base of each gastrostege, and usually a blackish blotch on each side of the center of each gastrostege. The urosteges are light gray outlined with blackish brown. Length to anus, 290 mm. Length of tail, 130 mm. Variation.—The principal variation in scale-characters is set forth in the following table. It will be noted that the Jervis and South Seymour snakes have three postoculars, while the Indefatigable specimen has only two. TABLE OF SCALE COUNTS, Dromicus steindachneri, new species e| 8 3s |e] 8 a |3 g 2|& 2 |8| 8 # | E\e 3 Ss 4 3 3 u te o 3) iS} 8 8 ra wn an oO (=) a A & aly = 10612 | ¢ | 19| 190] 114¢ | 1] 3-3 ole 8 | 1 | Jervis 10617 | ¢ |19|176| 97c¢ |1]3-3| 242 18] 1] Jervis | 10795), 3) 49 |/469:| /96ie) tell 2°p eae 1-8 it ose 10482 | 9 | 19] 176| 72 +|1| 3-3] 242 |8| 1 | Seymour sos | ¢ | 19] 176] 58+] 1] 3-3|{34? | 8] 4 | Seymour Vor. 1] VAN DENBURGH—SNAKES OF THE GALAPAGOS ISLANDS 355 The Jervis, Seymour and Indefatigable specimens all have light longitudinal stripes. The stripes are similar in position and color to those of Dromicus dorsalis. The whole coloration is so like that of striped specimens of D. dorsalis that the two species readily pass as one, until the scales are examined and the gastrosteges counted. On closer examination, however, one notes that in D. steindachneri the longitudinal light lines are broader, being three or four scales wide, each of the lateral scales has a central light area, and there usually is a blackish cross-bar at the base of each gastrostege, and often a blackish blotch on each side of the center of each gastrostege. The dorsal scales also sometimes have light centers. In the Jervis specimens the lower lateral scales are nearly as light as the light stripe. In No. 10617 a dark brown line runs along the lower border of the light stripe. The largest specimen measures 365 mm. from snout to vent. Habits —From the stomach of No. 10484 from South Sey- mour were taken the remains of a grasshopper. General remarks.—This interesting little snake is most closely related to Dromicus slevini. It is probable that both are either quite rare or very retiring in habits. It is a pleasure to associate with this handsome little species the name of Dr. Franz Steindachner, who was among the first to study the snakes of the Galapagos Archipelago. Hydrus platurus (Linnzus). Brcotor SEA-SNAKE No specimens of this snake have been taken in the Galapagos Archipelago, but the following note from Mr. Slevin’s diary shows that it occurs there.“ “Feb. 24, 1906. Sailed [from Chatham] for Hood Island. This afternoon at 4:15, Stewart sighted a sea-snake. King also saw it, and the boat was put out immediately, but we failed to get it, as it went under. King said it was about twenty inches long, black on the top and bright yellow below. We had some headway on, so passed it fairly quickly. This is the first one seen. Weather is very hot now and has been for the last few days. Light winds and strong currents make it hard to get around, and we have not made much progress during the day. Barrington, Chatham, Hood and Charles are in sight.” January 15, 1912 356 CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. EXPLANATION OF PLATE XXII Chart of the gastrostege counts in specimens of Dromicus. Dots indi- cate counts on specimens in the Academy and Stanford collections. Dots above the line are males, those below, females. Crosses indicate records from specimens not examined by me. ’ I DENBURG] PLATE XXil 237 230 « } 23 B9p 249 HL 24 A¢6 AHS 350 JSD [0 [oe ee ee ee Th eS eee ee Proc CatAcan. Sci 4.7 SER VoL.1 Chile and Peru ae i} Wy he 4 La PIV sb io uh "a Nice yea : ENG) a Ah bat 358 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4r Ser. EXPLANATION OF PLATE XXIII Dromicus biserialis (Gunther). No. 9448. Gardner Island, near Charles Island. Female. Tei At! Ty Me at aaah Tee tiene rhe ere paca i Me oi ! a iy Nit vy, 360 CALIFORNIA ACADEMY OF SCIENCES Proc. 4rm Sen. EXPLANATION OF PLATE XXIV Dromicus hoodensis new species pee . No, 11799. Type. Hood Island. Male. I b-] i ‘ f a } tat Proc CaLAcan. Scr 4 as, iit a aye Dy) Aten he aah aE ) oy a PON Ls oO OE LRAT TT an Rn evcishits Trey Laan PROT he vu ANl aia ailygt! Be e] enn a Ate NAD ny | IAN Wit : ee a ae Be Te ove) Aone BART of] Lay U) Hi Sy MOR ee cutt, Hy : ¥ YAS 362 ’ CALIFORNIA ACADEMY OF SCIENCES [Proc. 47H Ser. EXPLANATION OF PLATE XXV Dromicus dorsalis (Steindachner) No. 10303. Indefatigable Island. Male. Striped. No. 10233. Indefatigable Island. Female. Spotted. ural a ae 364 CALIFORNIA ACADEMY OF SCIENCES (Proc. 4TH Ser. EXPLANATION OF PLATE XXVI Dromicus dorsalis (Steindachner) No. 10183. Barrington Island. Male. Striped. No. 12061. Barrington Island. Male. Spotted. 4.™ Spr VoL I IDI C CAL.ACAU Prot a Milt Bren er | Penis | ae Ae ; ae ere OS oad ih ‘ - aon) ray Ace Te hDA aan aC ane Teel Wee i ethan DA rave are Miia te Ney) Pier lh! pen erat ii ne nf ar ae + (a pan | lye enh Na Uns rt minus ve fi in ; 366 CALIFORNIA ACADEMY OF SCIENCES EXPLANATION OF PLATE XXVII Dromicus occidentalis new species No. 11488. Type. Narborough Island. Female. 4) . Proc CaLAcap Sci 47! SeR VoL. aN ype a nar PODS! i ¥ Apia bal one Phe lca pein aici) Malia al Way iy MMOH i RASA i SUK aay Oe rf n ; she 7 n f ia u Miata a 7 UJ i a 1 qi a vP e - h d 7 368 CALIFORNIA ACADEMY OF SCIENCES — ‘[Paoc. 4rm Ser. a EXPLANATION OF PLATE XXVIII i ttit«*s Dromicus occidentalis helleri new subspecies . No. 10280. Type. Brattle Island. Male. CoN aay 370 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. EXPLANATION OF PLATE XXIX Dromicus slevint new species No. 12216. Type. Duncan Island. Male. No. 12159. Cowley Mt., Albemarle Island. Female. Proc CaLAcan. Scr 4.78 SER VoL 1 Piss Wa eae Vary ye ee An ip hy a ae ntact a AURiaS? Wi ha be ies ene Han Ru i 372 CALIFORNIA ACADEMY OF SCIENCES [P EXPLANATION OF PLATE XXX Dromicus steindachneri new species No. 10795. Type. Indefatigable Island. Male. No. 10484. South Seymour Island. Female. Proc CaLAcap Scr 47 Ser Vow | PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FourTH SERIES VoL. I, pp. 375-404 January 19, 1912 EXPEDITION OF THE CALIFORNIA ACADEMY OF SCIENCES TO THE GALAPAGOS ISLANDS, 1905-1906 Vv. NOTES ON THE BOTANY OF COCOS ISLAND BY ALBAN STEWART Botanist to the Galapagos Expedition During the autumn of 1905, while acting as Botanist of the scientific expedition sent to the Galapagos Islands by the Cali- fornia Academy of Sciences, our party stopped at Cocos Island from September 3rd to 13th inclusive, during which time a considerable collection of plants was made. Owing to the fact that Dr. B. L. Robinson of the Gray Herbarium, and Professor H. Pittier of the United States Department of Agriculture, have a catalogue of the plants of this island in preparation, in which all of the scattered refer- ences to its flora will be brought together, the present paper will deal mainly with the collection of plants and notes made by the author, so as not to infringe upon the work already done by these gentlemen. The collection was identified at the Gray Herbarium of Harvard University some three years ago, but owing to the more pressing need of getting the results of the work done in the Galapagos Islands ready for publication, the author has January 16, 1912 376 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. not been able to take up the less important results of the expedition until the present time. The work of identification was greatly facilitated through the kindness of Dr. Robinson in allowing me to use the list of plants already compiled by him. I wish here to express my thanks to Dr. Robinson for this, as well as for his kindness in allowing me to publish his description of Cecropia Pittieri, a new species of this genus which occurs abundantly on this island. I wish also to acknowledge my indebtedness to Dr. W. G. Farlow for identi- fication of the mosses; to Professor M. L. Fernald for much assistance; to Miss Mary A. Day, Librarian of the Gray Herbarium, for help in looking up the literature in connection with the subject; and to Mr. H. H. Bartlett of the United States Department of Agriculture, for identifying the speci- mens of Hypolytrum nicaraguense. The photographs were made by Messrs. R. H. Beck and E. W. Gifford, members of the expedition. Cocos is a small island which lies in longitude 86° 59’ 17” W., latitude 5° 32’ 57” N., and is about 300 miles distant from Costa Rica, to which country the island belongs. According to the chart issued by the Hydrographic Office, it is about 3% miles long in a north and south direction, 33g miles broad east and west, and rises to a height of 2788 feet. There are several small islets a short distance off shore, beyond which the water rapidly deepens, so that the thousand-fathom line is reached only a short distance away. There are only two places where an anchorage can be effected, and the interior of the island can be reached with safety. Chatham Bay, which lies on the north side, affords the best anchorage for vessels; since the waters are more quiet on this side of the island, and the sand beach at the end of the bay affords a good landing-place for boats. A small stream of water enters at the head of the bay, and, from the different dates cut in the rocks about the mouth of this stream, one would judge that it was often visited by vessels during the early part of the last century. On either side of this bay, east and west, there are tall cliffs heavily covered with tropical vegetation. Wafer Bay, on the northwest side of the island, is more exposed, and is subject at times to heavy swells which render anchorage less safe there than in Chatham Bay. With Vor. I] STEWART—BOTANY OF COCOS ISLAND 377 the exception of these two places and Dampier Head, on the southeast side of the island, the remainder of the shore is made up of tall cliffs, some of which must be a thousand or more feet in height, over the tops of which numerous waterfalls come tumbling down into the sea. Ten of these waterfalls were counted between Lionel Head and Berthaume Point, within a distance of less than a mile, while circumnavigating the island in a small boat. The settlement is located at Wafer Bay, where Captain August Gissler resides with his wife and a number of laborers. Several corrugated iron houses have been built there, and a small tract of land has been put under cultivation, in which a considerable number of domesticated plants and tropical fruits are grown. For some years past Captain Gissler has been in search of treasure which is supposed to have been buried on this island during the early part of the last century. Some portions of the treasure are reported to have been found. Captain Gissler is the duly appointed governor of the island, and is visited periodically by the Costa Rican gunboat to bring supplies and mail; but as the island lies out of the general track of both sailing vessels and steamers, it is seldom visited by other vessels. At the time of our visit the gunboat had not been out for some months, and in consequence some of the supplies had begun to run low. The sides of the mountain rise abruptly to a cone, which lies toward the west side of the island, about a mile and three- quarters from the settlement at Wafer Bay. Alternating ridges and deep canyons cover the mountain sides, rendering traveling almost impossible except along the tops of the ridges and along the beds of streams. According to Captain Gissler, a single ridge can be followed from the base to the top of the mountain, the ridges probably representing ancient lava-flows. Unfortunately none of the members of the party visited the top of the mountain, although an attempt was made to do so by following up an old trail. The trail had been made several years before, and as it had not been much used since, it had become heavily overgrown with vegetation, and could not be followed beyond 650 feet elevation. We could get very little information about the interior of the island, especially about the eastern part of it, which has never been visited so far as is 378 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. known. A fairly good view of this part of the island was obtained from the top of a cliff at Chatham Bay, and it appeared to be made up of a broad table-land heavily covered with vegetation, as are the remaining portions of the island. Rock-exposures occur along the banks of streams and along the sides of perpendicular cliffs. So far as could be observed, the rocks are basaltic in character. Columns of basalt occur frequently near sea-level, and caverns of considerable depth have been formed in many places by the action of the waves. The soil is composed for the most part of a sticky yellow clay and vegetable mold. From the more exposed places the mold has been washed off, leaving the clay bare. On the steep sides of the mountain erosion is rapid. In the small valleys one often encounters large forest trees which have been dislodged from the steep hillsides above by the washing away of soil from the roots to such an extent that they could no longer maintain their position. Land-slides are rather frequent, and when they occur, large quantities of earth and boulders are brought down along with the vegetation which covers the area. After a land-slide Jpomoca cathartica seems to be one of the first plants to invade the denuded area, followed by Hibiscus tiliaceus. The island lies in the moist tropical belt, and has a large amount of rainfall, the exact amount of which is not known, but it probably amounts to several feet per year. May, June, and July are said to be the rainiest months, and January, Feb- ruary, and March the driest. It rained eight out of the eleven days we were on the island, and some of the rains during this time were much harder than those which occur in more tem- perate regions. According to Captain Gissler the temperature ranges from 68° to 92° F. Halophytic plants are very few in number, possibly because of the precipitous nature of the shores in most places. Jpo- moea Pes-caprae is the most pronounced halophyte, and it occurs only to a limited extent on the sand beaches at Wafer Bay. Hibiscus tiliaceus forms small groves near the beach in a few places; and Clusia rosea often forms dense thickets along the sides of the cliffs some distance above the water, sending down absorbing roots into the sea. At several places near the shore there are small groves of Cocos nucifera, the nuts of Vou. 1] STEWART—BOTANY OF COCOS ISLAND 379 which are used to make oil for lighting purposes when the gunboat from Costa Rica delays its periodic trips too long. There are no mangroves, possibly because of the absence of quiet bays and lagoons. The interior of the island is covered for the most part with rain-forests, in which the vegetation is usually so dense that even at midday, with the sun shining, the light is almost as diffuse as at twilight. In such places there is an intense strug- gle among plants to gain the light—in consequence of which both epiphytes and lianes are very abundant in individuals, if not in species. The following list includes the species in the collection which are either epiphytes or lianes: Anthurium scandens Ipomoea cathartica Lycopodium linifolium Oleandra nodosa Philodendron sp. Selaginella Galeotti Tassadia colubrina Tillandsia sp. Trichomanes capillaceum. In addition to the above, there are several lianes which are in a sterile condition, so that even their generic relations cannot be determined. One of these is the most important liane on the island, extending in rope-like masses from tree to tree, often supporting hanging baskets of Tillandsias and other epiphytes. Unfortunately specimens of the large forest trees are but poorly represented in the collection, because of the fact that the most of the forest trees tower a hundred or more feet above the ground. Since the foliage is almost invariably at the top, specimens could not be obtained without cutting down the trees—which was too much of an undertaking. I used to look up longingly at the tops of these trees, wishing that I could obtain specimens; but I have since learned that it is the common experience of botanists to be unable to obtain speci- mens of the forest trees while collecting in tropical rain- forests. Besides the trees of Hibiscus and Clusia, mentioned above, there is at least one species of Cecropia which commonly 380 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. occurs along the faces of almost perpendicular cliffs, the roots being able to hold on to a mere crevice or shelf of rock, while the trunks grow up parallel with the wall of the cliff. There is also a species of palm which grows mostly above 400 feet elevation. Some specimens of this were collected lower down, but they are in a sterile condition. At least two species of Ficus occur here, one of which forms banyan trees of some size. The largest, and probably the most important tree from an economic standpoint, is one which bears the common name of “iron wood” according to Captain Gissler, who says that there are trees on the island so large that timbers 33X60 feet could be cut from them. The wood of this tree is dark brown in color and very hard. Underneath the trees there is usually a dense growth of bushes, so thick in most places that traveling through them is extremely difficult. In fact we found that the easiest way to get into the interior of the island was to follow up the beds of the larger streams, and occasionally make short excursions: off to the side. The most common bushes are; Eugenia pacifica, Clidemia hirta, C. umbonata, Miconia dodecandra, and Cli- badium acuminatum; three of which belong to Melastomaceae, and are the most abundant. Ferns also occur abundantly, forming a very important element of the undergrowth. Ex- tensive brakes are formed by Nephrolepis biserrata, especially where the large vegetation is more or less open. The moist banks along the sides of the streams are usually heavily covered with ferns, those which occur in such places being: Adiantum petiolatum, Asplenium cristatum, Ceropteris calo- melanos, Hymenophyllum sp., Polybotrya cervina, Polypo- dium aureum, Trichomanes crispum, and T. elegans. Also- phila armata is the only tree-fern found on the island. Filices are by far the largest family represented in the collection, twenty out of the seventy-seven species of vascular plants collected belonging to it. Of the remaining families of vascular plants there are none that contain more than five species, and the majority are represented by but one or two. Endemic species are included in the following: Chloris paniculata, Kyllinga nudiceps, Cecropia Pittieri, Eugenia pacifica, Osse@a macrophylla, Ardisia cuspidata, Bertiera angustifolia, and Clibadium acuminatum. Vor. I) STEWART—BOTANY OF COCOS ISLAND 381 On comparing the above with the number of endemic spe- cies found on the Galapagos Islands, one is at once struck with the small number of endemic species found on this island; and while the entire flora is not recorded in this paper, it is very likely that the number of species omitted is not large. It is of course unsafe to draw any very definite conclusions from incomplete data, yet it is safe to say that the per cent of endemic species on the Galapagos Islands is very much larger than on Cocos Island. It is interesting to note that but 8.697% of the species mentioned in this paper are endemic, while in the Galapagos Islands 40.9% are endemic. There is also an evident wide divergence in the total number of species found on the two, the Galapagos flora containing 682 species, while the Cocas flora very likely contains but little if at all over a hundred species. The wide divergence between the flora of the Galapagos Islands and that of Cocos Island, has been mentioned by authors who have written on these floras in the past. The following is a list of the species found on Cocos Island which are also found on the Galapagos Islands: Acrostichum aureum Adiantum petiolatum Asplenium cristatum Asplenium myriophyllum Dryopteris parasitica Nephrolepis biserrata Nephrolepis pectinata Polypodium aureum Polypodium lanceolatum Polypodium Phyllitides Polystichum adiantiforme Digitaria sanguinalis Eleusine indica Paspalum conjugatum Paspalum distichum Setaria setosa Commelina nudiflora Fleurya aestuans 382 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH SER. Anona cherimolia* Anona glabra Caesalpina bonducella Euphorbia pilulifera Ricinus communis* Hibiscus tiliaceus Ipomoea Bona-nox Ipomoea Pes-caprae Coffea arabica* From the above presentation it can be seen that the species common to the two groups of islands are for the most part those of rather wide distribution, and owing to the relatively small size of most of them, the general appearance and make-up of the two floras is but little influenced by them. The species which make up the bulk of the vegetation, especially the larger vegetation, are totally different on the two groups of islands—a fact which may have some significance. In a paper written some years ago by Dr. George Baur, an attempt was made to establish a former land-connection between the Galapagos Islands and the American continent, the connection presumably having been somewhere in the Mexican region. The improbability of such a connection has already been shown,t and it seems that the great difference in the floras of Cocos and the Galapagos islands strongly opposes Dr. Baur’s view. If there has ever been a land-mass connecting the Galapagos Islands with the mainland of North America, it must evidently have included the Cocos Island region, since its position is such that no considerable land-mass could have existed in this part of the ocean without including it. While the climatic conditions on the lower parts of the islands of the Galapagos group are entirely different from that of Cocos Island, being dry in one and moist in the other, the middle and upper por- tions of the higher islands of the Galapagos are moist, and capable, in places at least, of supporting fully as mesophytic vegetation as is Cocos—a fact which is evinced by the pres- ence of eleven ferns common to the two. A former land- * Probably introduced through cultivation into both the Galapagos Archipelago and Cocos Island. + American Naturalist, v. 25, 310 (1991). t Stewart. Proc. Calif. Acad. Sci. 4th Ser. v. 1, No. 2, pp. 233-239. Vor. I] STEWART—BOTANY OF COCOS ISLAND 383 connection between the two groups of islands should have left a much larger number of species common to the two than is actually found. The flora of Cocos, like that of the Galapagos Islands, is distinctly that of an oceanic island. The relatively large num- ber of ferns, the much smaller number of species in the remaining families, and the total number of species found on the island lend support to this view. The flora is probably of much more recent origin than is that of the Galapagos Islands. While the island lies nearer to the mainland by nearly three hundred miles, where presumably the various agents that dis- seminate seeds would work to at least as good advantage as in the Galapagos Islands, yet the number of species represented is probably not more than one-sixth as great. It seems pos- sible that the time that has elapsed since conditions on the island were suitable for the growth of higher vegetation has not been sufficient to stock the island by the slow process of seed dissemination, over considerable areas of water, with as many species as it is capable of supporting. The small number of endemic species on the island might also point to a relatively recent origin of its flora. The following are the species collected on the island by the author : FILICES Acrostichum L. A. aureum L. Sp. Pl. 1069 (1753): very abundant along the stream leading into Wafer Bay and on the hillsides up to 125 ft. It grows in large bunches 6-8 ft. high and with 30 or more fronds to a bunch, (No. 225). Further distr. general in tropical regions. Adiantum L. A. petiolatum Desv. Berl. Mag. V. 326 (1811): in crevices or rocks on the banks of the stream leading into Wafer Bay, (No. 226). Further distr. Mex., W. Ind., S. Am. Alsophila R. Br. A. armata (Sw.) Pr. Tent. 62 (1836). Polypodium arma- tum Sw. Prod. 134 (1788): very abundant on the banks of 384 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. the streams and on the hillsides surrounding both Chatham and Wafer Bays. It forms trees 8-15 ft. in height, and is apparently the only tree-fern on the island, (No. 227). Fur- ther distr. Mex., W. Ind., S. Am. Asplenium L. A. cristatum Lam. Encycl. IT. 310 (1786) : common on wet rocks on the side of a perpendicular cliff near Chatham Bay, (No. 228). Further distr. Mex., W. Ind., S. Am., Old World. Ceropteris Link. C. calomelanos (L.) Und. Bull. Torr. Cl. XXIX. 632 (1902). We 0 6 e'ee © vloleieeeivieicceree Slee jeter aisisrave ‘ds s